Part G - Discipline SoTA Patterns Kit
G.Core - Discipline SoTA Kit Invariants, Defaults and Refresh Triggers (Part G)
Status: Stable
Tag. Architectural pattern for shared Part G rules Stage. design‑time: authoring and stable citation Primary hooks. E.8 (pattern template), E.10 (lexical/ontological rules), E.19 (conformance discipline), A.6.7 (SuiteObligations + suite protocol pins), A.15.2 (edition/reference baseline), A.15.3 (typed planned filling when applicable), A.19.CN (CN‑Spec), G.0 (CG‑Spec), A.19.CHR (CHR suite boundary), C.23 (SoS‑LOG), F.17 (UTS), F.15 (unification SCR/RSCR).
Normativity. Normative unless explicitly marked informative
Purpose. Keep Part G invariants tied to their governing definitions, with stable references for defaults and refresh triggers. Use the RSCR trigger kind catalogue and Default Governing Definition Index below when authoring or changing Part G patterns.
Pattern placement. New Part‑G PatternIds are permitted when (i) they introduce a new kit/pack class (typically in G.2–G.5), or (ii) they are required to preserve one governing pattern per wiring extension and wiring-only separation. Method/discipline/generator specifics SHOULD default to GPatternExtension modules under G.x:Extensions (scoped by PatternScopeId = G.x:Ext.* and GoverningPatternId), rather than adding new Part‑G patterns.
G.Core:1 - Problem frame
Part G contains patterns for CG‑frame characterization and its downstream artefacts (cards, evidence graphs, bridge surfaces, refresh/shipping orchestration, parity harnesses, dashboards, interop surfaces). These patterns reuse invariants defined by their governing patterns as suite obligations and protocol norms.
Part‑G‑wide invariants are governed by G.Core so every G.x can:
- cite the core invariants rather than restating them, and
- place pattern-scoped specifics in
Extensions, keeping universal rules, kit surfaces, and descriptions of methods or generators separate.
G.Core:2 - Problem
Restating shared invariants in several Part G patterns can produce the following inconsistencies:
- Shadow governing specs emerge: downstream patterns restate CN‑Spec / CG‑Spec constraints, accidentally creating “local specs” that can diverge from the canonical governing definitions.
- Crossing discipline becomes inconsistent: “crossing events” and “crossing visibility” are described differently across
G.x, causing ambiguity about what must be pinned (UTS/Path/policy‑ids/editions) and what triggers refresh/regression. - Guard semantics drift: tri‑state eligibility and “unknown handling” can be reinterpreted in local prose, producing hidden fourth statuses or implicit coercions.
- Hidden scalarization. A pattern may leave a change undeclared when it collapses a partial order into a scalar or introduces a total order through a numeric summary.
- Suite/kit/pack mixing blurs governing-definition assignment: downstream patterns drift into “governing” what should remain governed by the suite boundary (
A.6.7andA.19.CHR), kit surfaces (eachG.x), or shipping (G.10). - Refactoring breaks public IDs: CC items and trigger labels become hard to evolve because removing duplicates risks breaking external references.
Part G requires a single place where these invariants and refactoring disciplines live, while keeping Part G patterns modular and method/discipline specifics explicitly separated.
G.Core:3 - Forces
- One governing definition vs. usability: We must centralize universal invariants, but
G.xmust remain readable and pattern-scoped for authors. - Delegation-first vs. completeness: Many norms already have canonical governing definitions such as
A.6.7,A.15.3,A.19,G.0,A.19.CHR, and the relevant Part E patterns.G.Coremust cite those governing definitions rather than duplicating semantics. - Public-id and alias continuity: Public CC IDs and deprecated trigger labels must remain stable as labels; deduplication must not break citations.
- Typed change control: RSCR/refresh must become id‑based (catalogued trigger kinds) rather than prose-based “meaning”.
- Strict distinction: Keep governing spec refs (CN‑Spec, CG‑Spec), suites, kits/surfaces, policies, planned baselines, audits, and refresh orchestration distinct.
- Minimal specificity naming: New IDs must be kind‑suffixed and minimally specific, to reduce semantic lock‑in while remaining precise.
- Scope discipline:
G.Coremust not become a container for discipline/method/generator taxonomies; those remain pattern-scoped (Extensions) or delegated to their governing patterns.
G.Core:4 - Solution
G.Core establishes Part‑G‑wide invariants as delegation rules + typed catalogs + authoring discipline.
G.Core:4.1 - Delegation-first citation for Part‑G‑wide invariants
For each Part G invariant already defined elsewhere, G.Core:
- standardises naming via
SuiteObligations.*(A.6.7:4.2), and - records where the invariant is governed, so downstream patterns cite rather than restate.
Governing definitions (normative index).
| Obligation handle | Canonical governing definition(s) | Part‑G note |
|---|---|---|
transport_declarative_only + cg_spec_cite_required_for_numeric_ops | A.6.7 + A.19.CN (CN‑Spec) + G.0 (CG‑Spec) + A.19.CHR | Cite the pinned edition of each specification that governs this use (CN‑Spec, CG‑Spec or both). Keep its definition in the governing pattern. |
semantic_correspondence_requirements | A.6.7 + F.9 + C.3.3 + E.18 | A claimed semantic correspondence between different local interpretation bases uses F.9; a kind correspondence uses C.3.3; a plane relation keeps its own predicate. The receiving use, reliance and any authorization remain separate. G.7 calibrates these claims; no CL minimum grants the use. |
crossing_visibility_required | E.18 (CrossingBundle) + A.6.7 | When a named downstream use requires durable E.18 crossing evidence, publish the required CrossingBundle. edition_key changes on crossing‑relevant artefacts (Bridge/CL surfaces, BridgeCards, CrossingBundle registries, and UTS rows for crossing artefacts) are treated as crossing-bundle edits. If the bundle required by that use is missing/non‑conformant, downstream consumers MUST abstain from that cross-Context or cross-plane use (no silent crossings). |
independent_correspondence_requirements | A.6.7 + C.3.3 | When a use actually relies on correspondence between distinct kinds, cite the obtaining KindBridge and separately check receiving admissibility and classification. A changed entity alone establishes no kind or semantic Bridge. |
| `guard_decision_tristate(pass | degrade | abstain)+unknown_never_coerces_to_pass` |
penalties_route_to_r_eff_only | A.6.7:4.2(2) | Model-bound penalty condition: apply the named model’s result meaning and applicability; loss assessment and numerical assurance penalties follow that governing clause. |
no_silent_scalarisation_of_partial_orders + no_silent_totalisation | A.6.7 | Preserve the declared result of the operation. A completed A.19.CPM comparison returns relation or poset tokens. A completed G.5-3 Select returns a set of candidates. G.5-6 DeclareSetResult identifies members already fixed for a use, including framework editions. A scalar ranking or total order needs an admitted comparator or policy for this use (CC‑GCORE‑SET‑1). |
planned_filling_in_workplan + actual_launch_facts_and_witness + gate_decision_separation | A.15.2 + A.15.3 + A.19.CHR + A.6.7 + E.18:5.6 | A baseline adopted as planned work is A.15.2 WorkPlan content; A.15.3 applies only to typed planned filling of an independently declared argument or relation position. Preserve the planned values. An actual launch claim requires independently admitted Work and obtaining direct relations or exact application bindings; any FinalizeLaunchValues witness is a separate episteme. GateDecision/DecisionLog semantics remain with their gate governor. |
DefaultGoverningDefinitionIndex.single_governing_definition_per_DefaultId | this pattern | Any default names exactly one governing definition; G.Core.DefaultGoverningDefinitionIndex is an index, not a second spec. |
This pattern also governs four pieces of Part‑G‑wide infrastructure that are not already governed elsewhere:
- the typed RSCRTriggerKindId catalogue (single writer),
- the Default Governing Definition Index (one governing definition per DefaultId; index only), and
- the Δ‑discipline for ID‑stable deduplication (delegation without public‑ID breakage), and
- the linkage compression catalogues (
GCoreConformanceProfileId,GCoreTriggerSetId,GCorePinSetId) used to keepG.xlinkage sections small.
G.Core:4.2 - State the applicable G.Core linkage
Each Part-G pattern makes its applicable core obligations, pins, defaults and refresh triggers recoverable. A direct reference with its use conditions is sufficient; GCoreLinkageManifest provides a compact representation when several shared sets are used.
-
Cite
G.Coreand the governing definitions of the obligations used. -
When using a
GCoreLinkageManifest, identify the applicable sets or individual ids:CoreConformanceProfileIds := { GCoreConformanceProfileId… }and/orCoreConformanceIds := { CC‑GCORE‑… }RSCRTriggerSetIds := { GCoreTriggerSetId… }and/orRSCRTriggerKindIds := { RSCRTriggerKindId… }CorePinSetIds := { GCorePinSetId… }and/orCorePinsRequired := { … }(list additions or overrides when pin sets are used)DefaultsConsumed := { DefaultId… }(the governing definition is resolved throughG.Core.DefaultGoverningDefinitionIndex)TriggerAliasMapRef?when local trigger tokens need canonical aliases
Nil‑elision (normative size rule). Any field whose value is ∅ MAY be omitted; omission means ∅ and does not relax any obligation.
Expansion rule (normative). If profile/set ids are used, the effective CoreConformanceIds / RSCRTriggerKindIds / CorePinsRequired are the unions of their expansions plus any explicitly listed ids (see G.Core:4.2.2, G.Core:4.2.3, and G.Core:4.3.4.2).
G.Core:4.2.1 - GCoreLinkageManifest (canonical shape)
GCoreLinkageManifest groups the pattern-local references to G.Core without duplicating the governing rules.
A G.x MAY render this linkage as prose, a table, or structured notation. Its applicable ids and conditions SHALL remain recoverable:
GCoreLinkageManifest := ⟨ CoreConformanceProfileIds?: {GCoreConformanceProfileId…}, CoreConformanceIds?: {CC‑GCORE‑…}, RSCRTriggerSetIds?: {GCoreTriggerSetId…}, RSCRTriggerKindIds?: {RSCRTriggerKindId…}, CorePinSetIds?: {GCorePinSetId…}, CorePinsRequired?: {…pin ids…}, DefaultsConsumed?: {DefaultId…}, TriggerAliasMapRef?: TriggerAliasMapRef ⟩
G.Core:4.2.2 - GCoreConformanceProfileId catalogue (compression primitive)
A GCoreConformanceProfileId is a stable identifier for a named set of CC‑GCORE‑* items. Citing it in a G.x linkage section is shorthand for listing those items.
| GCoreConformanceProfileId | Expands to CC‑GCORE‑* (set) | Notes |
|---|---|---|
GCoreConformanceProfileId.PartG.AuthoringBase | {CC‑GCORE‑CN‑CG‑1, CC‑GCORE‑CROSS‑1, CC‑GCORE‑PEN‑1, CC‑GCORE‑SET‑1, CC‑GCORE‑P2W‑1, CC‑GCORE‑DEF‑1, CC‑GCORE‑TRIG‑1, CC‑GCORE‑TRIG‑2, CC‑GCORE‑TRIG‑3, CC‑GCORE‑TRIG‑4, CC‑GCORE‑ID‑1, CC‑GCORE‑ID‑2, CC‑GCORE‑LINK‑1, CC‑GCORE‑LINK‑2} | Default baseline for most Part‑G kits. |
GCoreConformanceProfileId.PartG.TriStateGuard | {CC‑GCORE‑GUARD‑1} | Add when the kit defines/consumes eligibility/guard outcomes. |
GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted | {CC‑GCORE‑UTS‑1} | Add when the kit mints/evolves public ids (UTS rows). |
GCoreConformanceProfileId.PartG.ShippingBoundary | {CC‑GCORE‑SKP‑1} | Add when shipping boundaries are in scope for the kit. |
G.Core:4.2.3 - GCorePinSetId catalogue (compression primitive)
A GCorePinSetId is a stable identifier for a named set of pin obligations that recur in Part‑G kits. Citing it in a G.x linkage section is shorthand for listing the member obligations.
Conditional pins (normative). A pin marked with ? is required when the pattern’s actual object or receiving use calls for that surface; otherwise it may be omitted. An independently selected kit or receiving contract can require a pin explicitly. Selection of a shared pin set preserves those conditions; it does not create a CG-Frame, comparison specification, WorkPlan or crossing.
| GCorePinSetId | Expands to CorePinsRequired (set) | Notes |
|---|---|---|
GCorePinSetId.PartG.AuthoringMinimal | {CG-FrameContext?, entityOfConcern?, GroundingHolon?, ReferencePlane?, CNSpecRef.edition?, CGSpecRef.edition?} | Identify what each claim concerns, including the compared targets when its rule calls for them. Name its empirical ground, plane and specification editions as required by the actual kit. A single reference may fill two roles when the entity is the same. Retain required CN/CG pins in a governed comparison. For an earlier kit carrying only ground and plane, use its source claim and frame when they identify the subject uniquely. Otherwise preserve those references and recover the missing subject before using its result. |
GCorePinSetId.PartG.CrossingVisibilityPins | {BridgeId?/BridgeCardId?, KindBridgeAssertionRef[]?, PlaneRelationRef[]?, BridgeMatrixId?, CL?/CL^k?/CL^plane?, Φ/Ψ/Φ_plane policy-ids?, CrossingBundleId?, UTSRowId[]?, PathId[]?/PathSliceId[]?} | Activate the channel and receiving-use conditions below. Selecting this pin set alone creates no correspondence, calibration, loss model, publication row or flow/gate crossing. |
CrossingVisibilityPins activation. Recover the relation and receiving use before expanding its applicable pins:
| Actual claim or use | Required references |
|---|---|
| F.9 semantic correspondence | the obtaining Bridge and its BridgeCard/source basis required by F.9; CL when the use reports or consumes that calibration |
| C.3.3 kind correspondence | exact kind endpoints and KindBridge assertion/basis; CL^k when its governing calibration or receiving rule uses it |
| Separately governed plane relation | its exact relation reference, source/target planes and direct rule; CL^plane only for a defined calibration used here |
| Numerical loss or assurance calculation | the applicable domain model and policy actually used; include Φ, Ψ or Φ_plane references when the model uses them. Any numerical assurance penalty satisfies A.6.7:4.2(2). |
| E.18 flow crossing, A.21 gate, public naming or path-citable reliance | the bundle, gate, UTS and path anchors required by that independently applicable rule; an edition change alone supplies none of these objects |
When more than one channel is used, retain every channel’s independently required basis. A no-crossing use has no crossing pins; a same-kind new entity or changed source edition alone creates no Bridge. The conditional marks permit omission only for an inactive condition, not omission of a reference required by a real calibration, assurance or gate claim.
G.Core:4.3 - RSCR Trigger Catalogue and docking discipline
G.Core defines the canonical Part G trigger kinds.
G.Core:4.3.1 - Definitions
-
RSCRTriggerKindId Canonical, stable identifier for a trigger kind (a class of “why RSCR/refresh must fire”). Cross-pattern reason code.
-
RSCRTriggerAliasId Pattern-scoped human label/token kept for ergonomics and alias continuity (e.g.,
G.11:T4,G.6:H3:lane-tag correction). -
TriggerAliasMap Mapping table:
RSCRTriggerAliasId → {RSCRTriggerKindId…}(1..n). -
RSCRTrigger Minimal conceptual form (notation-independent):
RSCRTrigger := ⟨ triggerKindId: RSCRTriggerKindId, scope: PathSliceId[] | PathId[] | PatternScopeId, payloadPins: { …id pins… } ⟩Include in
payloadPinsthe pins needed to act on the trigger, such as edition pins, policy IDs, Bridge IDs, evidence pins, and regression-set IDs.
G.Core:4.3.2 - Governing-definition model
- TriggerGoverningDefinition :=
G.Core. - Any new trigger kind SHALL be added to
G.Corefirst. - Other patterns MAY define aliases only (or cite shared alias maps), and MUST map aliases to canonical kinds.
G.Core:4.3.3 - Authoring rules
-
No implicit triggers: Any RSCR/SCR/refresh artefact that records reasons MUST identify them by their canonical
RSCRTriggerKindIdvalues. It may also record aliases as labels. -
No implicit overloading: A local token string (e.g.,
T4) SHALL NOT silently change meaning across patterns; namespace is part of the alias (G.11:T4≠A.20:T4). -
Granularity discipline: If a local cause is narrower than an existing canonical kind, map it to that kind and keep the nuance as a local scope note. If the difference matters for planning/selection, add a new canonical kind.
-
Multi-cause discipline: When an event spans multiple canonical kinds, record multiple triggers (preferred) or map the alias to a set
{…}and require emitting the full set.
G.Core:4.3.4 - Canonical trigger catalogue
Use the canonical kinds defined below to classify refresh causes and interpret the deprecated G.6:H3 and G.11:T0…T7 aliases used across G.0…G.13.
G.Core:4.3.4.1 - Canonical kind definitions (normative, minimal)
Each RSCRTriggerKindId SHALL have a short, stable definition in G.Core (single-writer) to prevent semantic drift.
| RSCRTriggerKindId | Minimal meaning (cause class) | Typical payload pins (non-exhaustive) |
|---|---|---|
RSCRTriggerKindId.LegalitySurfaceEdit | A legality surface changed (CG‑Spec: ComparatorSet/SCP/Γ_fold/MinimalEvidence, or equivalent legality inputs). | CGSpecRef.edition, ComparatorSetRef.edition, SCPRef.edition, ΓFoldRef.edition |
RSCRTriggerKindId.PenaltyPolicyEdit | A penalty / Φ / Ψ / FailureBehavior / SoS‑LOG branch policy changed. | penalty policy ids, Φ/Ψ policy ids, SoS‑LOG branch id pins |
RSCRTriggerKindId.CrossingBundleEdit | A crossing bundle changed (Bridge/CL routing, crossing-bundle registry cards, crossing policy pins), including edition_key changes of crossing‑relevant artefacts (BridgeCards, CrossingBundle registries, UTS rows for crossing artefacts). | BridgeId/BridgeCardId, BridgeMatrixId?, CL* ids, crossing policy ids, UTSRowId[], PathId/PathSliceId? |
RSCRTriggerKindId.ReferencePlaneEdit | ReferencePlane or plane-routing surface changed. | ReferencePlaneId, plane-policy ids |
RSCRTriggerKindId.EditionPinChange | Any pinned edition relevant to downstream artifacts changed (including CNSpecRef.edition, CGSpecRef.edition, comparator/method/descriptor/distance/etc.). Crossing‑artefact edition_key changes are additionally classified as CrossingBundleEdit per multi‑cause discipline. | changed *.edition pins, affected PathSliceIds |
RSCRTriggerKindId.TokenizationOrNameChange | A published tokenization / naming / alias surface changed in a way that can affect docking, citations, or dispatch (e.g., UTS Name Cards, twin labels, alias maps). | affected UTSRowId[], NameCardId[], alias ids / maps |
RSCRTriggerKindId.PolicyPinChange | A policy-id pin used by characterization changed (selection, insertion, emission, routing, refresh policy, etc.). | policy ids (and other non-edition configuration pins when they are explicitly pinned) |
RSCRTriggerKindId.TelemetryDelta | A telemetry input that influences a refresh or selection decision changed. Display-only updates fall outside this trigger kind. | telemetry ids/refs, Audit-published pins |
RSCRTriggerKindId.FreshnessOrDecayEvent | Time/freshness/decay conditions affecting validity changed (window shift, decay thresholds, freshness policy edits). | freshness window refs/ids, decay/freshness policy ids |
RSCRTriggerKindId.EvidenceSurfaceEdit | Evidence graph / evidence surface changed in ways that affect admissibility/acceptance/comparison. | evidence pins, EvidenceGraph refs, affected PathIds |
RSCRTriggerKindId.MaturityRungChange | Maturity rung/ladder state changed for relevant artifacts or paths. | maturity rung ids, affected scopes |
RSCRTriggerKindId.BaselineBindingEdit | Selected baseline references or applicable typed fillings changed, opening the affected P2W-use question. | Exact WorkPlan and baseline locator; declared-position row designators only when typed filling applies; changed pins and relevant variance |
RSCRTriggerKindId.DefaultGoverningDefinitionChange | The governing definition of a DefaultId (as recorded in G.Core.DefaultGoverningDefinitionIndex) changed, or a default row was added, deprecated or removed. | affected DefaultId.*, previous and new governing definition references where each exists |
G.Core:4.3.4.2 - Canonical trigger sets (compression primitive)
GCoreTriggerSetId identifies a named set of RSCRTriggerKindId values. A G.x MAY cite trigger sets in RSCRTriggerSetIds instead of repeating long RSCRTriggerKindIds lists.
| GCoreTriggerSetId | RSCRTriggerKindIds (set) | Notes |
|---|---|---|
GCoreTriggerSetId.CGSpecGate | {RSCRTriggerKindId.LegalitySurfaceEdit, RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.ReferencePlaneEdit, RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.FreshnessOrDecayEvent} | Covers CG‑Spec legality‑gate kits (e.g., G.0). |
GCoreTriggerSetId.SoTAHarvestSynthesis | {RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.TokenizationOrNameChange, RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.ReferencePlaneEdit, RSCRTriggerKindId.LegalitySurfaceEdit, RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent} | Covers SoTA harvesting/synthesis packs (e.g., G.2). |
GCoreTriggerSetId.EvidenceGraphKit | {RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.PenaltyPolicyEdit, RSCRTriggerKindId.ReferencePlaneEdit, RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.FreshnessOrDecayEvent, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.MaturityRungChange, RSCRTriggerKindId.BaselineBindingEdit} | Covers EvidenceGraph/SCR kits (e.g., G.6). |
GCoreTriggerSetId.BridgeCalibrationKit | {RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.PenaltyPolicyEdit, RSCRTriggerKindId.ReferencePlaneEdit, RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.FreshnessOrDecayEvent, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.BaselineBindingEdit} | Covers bridge calibration/CL kits (e.g., G.7). |
GCoreTriggerSetId.RefreshOrchestration | {RSCRTriggerKindId.LegalitySurfaceEdit, RSCRTriggerKindId.PenaltyPolicyEdit, RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.ReferencePlaneEdit, RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent, RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.MaturityRungChange, RSCRTriggerKindId.BaselineBindingEdit} | Covers refresh orchestration (e.g., G.11). |
G.Core:4.3.5 - Initial alias maps
These alias maps are normative docking artefacts and preserve deprecated alias labels while moving semantics to canonical ids.
TriggerAliasMap.G11
Based on the existing trigger catalogue in G.11 (T0…T7).
G.11:T0 → { RSCRTriggerKindId.PolicyPinChange }G.11:T1 → { RSCRTriggerKindId.TelemetryDelta }G.11:T2 → { RSCRTriggerKindId.EditionPinChange }G.11:T3 → { RSCRTriggerKindId.EditionPinChange }G.11:T4 → { RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.PenaltyPolicyEdit }G.11:T5 → { RSCRTriggerKindId.FreshnessOrDecayEvent }G.11:T6 → { RSCRTriggerKindId.MaturityRungChange }G.11:T7 → { RSCRTriggerKindId.PolicyPinChange }
TriggerAliasMap.G0 (reserved).
Map any stable deprecated registry-hook labels emitted/recorded by G.0 to the canonical kinds above (typically LegalitySurfaceEdit, PenaltyPolicyEdit, CrossingBundleEdit, ReferencePlaneEdit, TokenizationOrNameChange), preserving the original label text as RSCRTriggerAliasId. If none exist, G.0 SHOULD emit canonical RSCRTriggerKindId values directly.
TriggerAliasMap.G6
EvidenceGraph H3 example causes → canonical kinds:
G.6:H3:freshness/decay change → { RSCRTriggerKindId.FreshnessOrDecayEvent }G.6:H3:Bridge CL/CL^k or loss update → { RSCRTriggerKindId.CrossingBundleEdit }G.6:H3:Φ/Ψ policy change → { RSCRTriggerKindId.PenaltyPolicyEdit }G.6:H3:lane tag correction → { RSCRTriggerKindId.EvidenceSurfaceEdit }G.6:H3:ReferencePlane correction → { RSCRTriggerKindId.ReferencePlaneEdit }G.6:H3:QD/OEE artefact updates (U.DescriptorMapRef.edition/DistanceDef, EmitterPolicyRef, InsertionPolicyRef, archive K-capacity) → { RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange }
G.Core:4.4 - Default Governing Definition Index
The index below identifies one governing definition per DefaultId: a CC item, policy-id or TaskSignature rule. It also states the default’s applicability conditions.
G.Core:4.4.1 - Definitions
-
DefaultId Stable identifier of a default (a default constant or default rule).
-
DefaultGoverningDefinitionRef A reference to the governing definition of the default (e.g., a CC item id like
CC‑G5.23, or a policy id, or a TaskSignature rule definition).
G.Core:4.4.2 - Rules
- Exactly one governing definition per
DefaultId. - Any other mention in
G.xMUST be a citation/delegation to the governing definition, not a competing statement. - A default may be conditional (default-rule) with explicit applicability conditions.
- The Default Governing Definition Index SHALL contain only defaults. Cite mandatory invariants through
CC‑GCORE‑…and their governing definitions.
G.Core:4.4.3 - Default governing definitions
| DefaultId | DefaultGoverningDefinitionRef | Notes |
|---|---|---|
DefaultId.PortfolioMode | CC‑G5.23 | Existing governing definition; other mentions delegate to it. |
DefaultId.DominanceRegime | CC‑G5.28 | Existing governing definition; other mentions delegate to it. |
DefaultId.GammaFoldForR_eff | CC‑G5.4 | Model-qualified support composition under B.3/C.2.2; no universal numeric fold. Keep separate support when no common model is justified. |
To add a default, name its governing definition. Use G.Core as that definition only for a Part G default not governed elsewhere. Any change in a row (addition, deprecation, removal or a change of governing definition) SHALL be treated as a refresh‑sensitive edit and recorded as RSCRTriggerKindId.DefaultGoverningDefinitionChange (payload: the affected DefaultId.* and the previous and new governing definition references where each exists).
G.Core:4.5 - ID continuity protocol (Δ‑discipline)
When moving universal norms out of G.x into G.Core:
- existing public CC ids in
G.xthat may be referenced externally SHALL NOT be deleted or renamed; - such CC items SHALL become delegation items that point to the relevant
CC‑GCORE‑…item(s); - each
G.xSHALL add exactly one bridge CC itemCC‑Gx‑CoreRef(first in its CC list) that makes linkedCC‑GCORE‑…items mandatory forG.xconformance.
Deprecated trigger labels (e.g., G.11:T*, G.6:H3:*) are preserved as aliases and MUST map to canonical trigger kinds.
The same Δ-discipline applies to every non-CC public identifier, including UTSRowId and RSCRTriggerAliasId. Publications MUST preserve earlier ids. Use aliases, deprecation notices and edition changes to express drift (see F.17 (UTS)). When downstream use may be affected, emit the applicable canonical trigger kinds (RSCRTriggerKindId.TokenizationOrNameChange, RSCRTriggerKindId.EditionPinChange).
G.Core:4.6 - Explicit non-goals
G.Core does not:
- introduce CG‑frame kit entities (e.g., BridgeMatrix/ReferencePlane/Φ registries); those remain in their governing
G.x; - introduce method-family taxonomies, discipline packs, or generator orchestration mechanisms; those remain as
Extensionsin their governing definitions (e.g., synthesis/shipping/refresh patterns); - define refresh algorithms; it defines trigger kinds and docking only.
G.Core:5 - Archetypal grounding
Tell.
A Part G pattern can present a short normative Core linkage section alongside its pattern-scoped Extensions. A direct reference with its use conditions also suffices (§4.2). Authors cite the governing definitions of universal obligations, including A.6.7, A.15.3, A.19, G.0, and A.19.CHR. Typed RSCR trigger kinds and DefaultGoverningDefinitionRef references point to named definitions.
Show 1: Refresh triggers without semantic drift.
G.11 uses trigger tokens T0…T7. G.Core keeps them as aliases and maps them to canonical trigger kinds (e.g., TelemetryDelta, EditionPinChange, CrossingBundleEdit). This makes RSCR reason codes consistent across patterns and avoids re-explaining trigger semantics in every pattern.
Show 2: Resolving competing defaults.
If several patterns imply a default for PortfolioMode, the Default Governing Definition Index points to the governing definition, CC‑G5.23. Other patterns, including bundle and log patterns, must leave the definition there and cite or delegate to it. A separate local wording could change how their readers interpret the default.
G.Core:6 - Bias-annotation (informative)
-
Centralization bias: One governing hub can become too thick. Mitigation: delegation-first citation; keep only Part‑G‑wide invariants and typed indices here.
-
Over-typing bias: A trigger catalogue can become overly granular. Mitigation: granularity discipline + scope notes; only add new kinds when planning/selection needs it.
-
Refactor rigidity bias: Preserving IDs can feel cumbersome. Mitigation: delegation items preserve IDs while enabling deduplication.
-
Default absolutism bias: Defaults may require conditional rules. Mitigation: Default Governing Definition Index allows conditional default rules with explicit applicability conditions.
-
Pragmatic bias: adds authoring overhead (linkage sections, alias maps). Mitigation: For ordinary linkage, cite the obligations with their use conditions directly or use a short linkage section. When moving shared norms from a
G.xintoG.Core, put exactly oneCC‑Gx‑CoreReffirst in the affected checklist (§4.5). -
Didactic bias: risks “glossy hub prose” that hides missing CC coverage. Mitigation: enforce CC/Solution coherence (E.19) and keep invariants checkable via
CC‑GCORE‑….
G.Core:7 - Conformance checklist (normative) — CC‑GCORE
The following conformance items retain their stated applicability conditions. Each Part-G pattern cites the items it uses, directly or through its linkage manifest.
| ConformanceId | Statement |
|---|---|
| CC‑GCORE‑DEL‑1 | A conforming G.Core SHALL be delegation-first: if a norm is already governed by A.6.7, A.15.3, A.19, G.0, A.19.CHR, or a relevant Part E pattern, G.Core cites that governing definition rather than duplicating semantics. |
| CC‑GCORE‑CN‑CG‑1 | A Part-G use governed by CN‑Spec or CG‑Spec SHALL cite the applicable specification and edition. No Part-G pattern introduces a shadow specification; apply SuiteObligations.transport_declarative_only and SuiteObligations.cg_spec_cite_required_for_numeric_ops under their governing conditions. |
| CC‑GCORE‑OBL‑1 | A conforming G.Core SHALL treat the obligation vocabulary in A.6.7:4.2 as the canonical naming surface for Part‑G‑wide obligations and SHALL NOT introduce competing obligation names for the same norms. |
| CC‑GCORE‑CROSS‑1 | A Part-G pattern SHALL apply A.6.7’s correspondence and visibility obligations only to the actual relations and receiving uses defined there. An F.9 correspondence, C.3.3 kind correspondence and separately governed plane relation keep their own bases; an actual E.18 flow crossing or A.21 gate keeps its required bundle/gate anchors. Changes to crossing-relevant artefacts trigger RSCRTriggerKindId.CrossingBundleEdit and, for a changed edition pin, EditionPinChange under multi-cause discipline. A changed entity, context/plane label or edition alone creates neither Bridge nor crossing. |
| CC‑GCORE‑GUARD‑1 | A Part‑G pattern SHALL treat `GuardDecision := {pass |
| CC‑GCORE‑PEN‑1 | A Part-G pattern SHALL apply A.6.7:4.2(2) to its loss and assurance-penalty claims (SuiteObligations.penalties_route_to_r_eff_only), including that clause’s model-bound applicability and separation from legality, invariant and claim-scope judgments. |
| CC‑GCORE‑SET‑1 | A Part‑G pattern SHALL preserve the result type of the admitted operation under a partial order and SHALL prohibit silent scalarization or totalization. A completed A.19.CPM Compare returns a set of relation or poset tokens in ComparisonResultSlot. CompareEligibility returns a separate guard decision. On abstention the comparison has no returned binding. A completed G.5-3 selection returns its candidate set. G.5-6 may declare exact members under an existing inclusion decision. An abstention or handoff has its own grounds. Any scalar summary SHALL remain report-only unless admitted as a comparator for this use (SuiteObligations.no_silent_scalarisation_of_partial_orders, SuiteObligations.no_silent_totalisation). |
| CC‑GCORE‑SKP‑1 | A Part‑G pattern SHALL preserve the suite/kit/pack distinction (A.19.CHR) and SHALL keep shipping concerns governed by their canonical governing patterns (e.g., G.10) rather than embedding shipping semantics into unrelated kits or core invariants. |
| CC‑GCORE‑P2W‑1 | When a Part-G use adopts an edition/reference baseline as planned work, identify its A.15.2 WorkPlan and local content locator. A.15.3 typed planned filling applies only to an independently declared argument or relation position, with its own meaning and binding rule. Final launch values and actual bindings remain enactment claims; gate decisions keep their own governor. Ordinary source or edition citation alone does not assert planned work. |
| CC‑GCORE‑LINK‑1 | Each G.x makes its applicable core obligations, pins, defaults and trigger kinds recoverable under G.Core:4.2, through direct references or a GCoreLinkageManifest. Nil-elision for ∅ does not waive an applicable obligation. |
| CC‑GCORE‑LINK‑2 | A CC‑Gx‑CoreRef, when used, delegates to the effective core obligations and their conditions, including referenced profile expansions. Equivalent direct delegation is sufficient. |
| CC‑GCORE‑UTS‑1 | If a Part‑G pattern mints, deprecates, or evolves any public identifier, it SHALL publish/update the corresponding UTS entries and cite them via UTSRowId pins, delegating UTS semantics (Tech and applicable Plain designations, alias/deprecation discipline, edition pins) to its canonical governing definition F.17 (UTS). |
| CC‑GCORE‑ID‑1 | When deduplicating, existing public CC ids in G.x SHALL NOT be deleted/renamed; they SHALL become delegation items to relevant CC‑GCORE‑… items. |
| CC‑GCORE‑ID‑2 | Publications SHALL preserve non-CC public ids, including UTSRowId and RSCRTriggerAliasId tokens (e.g., T0…T7). Express changes through aliases, deprecation notices and new editions. An old id SHALL NOT be reused or redefined. When downstream behaviour can change, the change SHALL emit a canonical RSCRTriggerKindId (e.g., TokenizationOrNameChange, EditionPinChange). |
| CC‑GCORE‑TRIG‑1 | A conforming G.Core SHALL define the canonical RSCRTriggerKindId catalogue and SHALL be its single writer. |
| CC‑GCORE‑TRIG‑2 | Any G.x that uses local trigger tokens SHALL provide (or cite) a TriggerAliasMap mapping each alias to canonical RSCRTriggerKindId. |
| CC‑GCORE‑TRIG‑3 | Any artefact that records RSCR/SCR/refresh reasons SHALL record canonical RSCRTriggerKindId (aliases may be recorded as labels only). |
| CC‑GCORE‑TRIG‑4 | A conforming G.Core SHALL keep TriggerAliasMap.* consistent with the governing patterns’ deprecated trigger alias catalogues (e.g., G.11:T*). Any change to an alias mapping SHALL be treated as refresh‑sensitive; at minimum it SHALL be recorded/emitted as RSCRTriggerKindId.TokenizationOrNameChange (and, if the mapped trigger kinds change, the corresponding canonical kinds apply as well). |
| CC‑GCORE‑DEF‑1 | A conforming G.Core SHALL maintain a Default Governing Definition Index for Part‑G defaults, ensuring each DefaultId.* names exactly one governing definition (a CC item, policy id or TaskSignature rule). All other patterns SHALL cite the governing definition and SHALL NOT state competing defaults. Any governing-definition change MUST be recorded as RSCRTriggerKindId.DefaultGoverningDefinitionChange. |
G.Core:8 - Common anti-patterns and how to avoid them
-
Anti-pattern: Restating CN‑Spec/CG‑Spec rules inside a
G.x“for convenience”. Avoid: cite the applicable specification (A.19.CN,G.0or both) throughCC‑GCORE‑CN‑CG‑1. -
Anti-pattern: Adding a fourth guard status (“unknown”, “maybe”, “probe-only”) as a separate decision value. Avoid: keep guard domain tri‑state; express “probe-only” as policy/branching and record via pins/audit.
-
Anti-pattern: Treating mandatory invariants as “defaults” to centralize them. Avoid: Cite mandatory invariants through their
CC‑GCORE‑*items and governing definitions. Reserve the Default Governing Definition Index for defaults, including constants and conditional default rules. -
Anti-pattern: Turning partial orders into scalar ranks silently. Avoid: report what the declared comparator established about its operands. To obtain a frontier, apply the dominance rule to the candidate domain (C.18). Use G.5-3 to select candidates or G.5-6 to declare members already fixed for this use. Any ranking that guides a decision needs an admitted comparator or policy.
-
Anti-pattern: Competing defaults scattered across multiple patterns. Avoid: Default Governing Definition Index; delegate duplicate statements to the one governing definition.
-
Anti-pattern: Local trigger tokens without canonical mapping. Avoid: provide/cite a
TriggerAliasMapwith namespace‑qualified aliases. -
Anti-pattern: Breaking public CC ids during dedup. Avoid: convert to delegation items; preserve IDs.
G.Core:9 - Consequences
- Positive: Part‑G patterns cite
G.Coreto find the governing definitions of shared invariants; refactors become safer and easier to audit. - Positive: RSCR becomes reason-code driven (typed triggers), improving traceability and preventing semantic drift.
- Positive: Default conflicts become detectable and resolvable because each
DefaultIdnames one governing definition. - Negative: Authors must identify the core obligations their
G.xuses. A dedicated linkage section adds work when that form is chosen. Moving shared norms intoG.Coreentails the checklist changes in §4.5. - Negative: Requires careful governance of the trigger catalogue to avoid excessive fragmentation.
G.Core:10 - Rationale
Citing one governing definition lets authors change an invariant without maintaining competing statements across Part G. Stable trigger kinds make refresh causes comparable across patterns. The Default Governing Definition Index lets an author resolve a default to its rule and applicability conditions.
G.Core:11 - SoTA alignment (informative)
The following approaches inform Part G authoring:
- Selective prediction / abstention informs tri‑state guard discipline: abstaining or degrading is a first-class outcome, not an error coerced into a scalar.
- Set-valued / conformal methods motivate set-return semantics: when comparability is partial or uncertainty is structural, sets or regions can preserve alternatives that a single answer would conceal.
- Multiobjective optimization and quality-diversity reinforce declared set-result and
Archivesemantics instead of forced “best single scalar”. - Monotone constrained modelling (where used) supports “legality-first” scoring/aggregation: constraints and admissibility precede optimization, mirroring CG‑Spec gate discipline.
- Schema evolution and contract testing motivate id-stable conformance points and typed trigger catalogues: stable identifiers + regression hooks are the practical mechanism for safe refactoring.
G.Core:12 - Relations
-
Builds on:
E.8pattern template and section disciplineE.10lexical/ontological rules (strict distinction; twin naming; kind‑suffix discipline)E.18CrossingBundle (crossing visibility bundle)E.19conformance disciplineA.6.7SuiteObligations + suite protocol pins (delegation support)A.15.2WorkPlan content for the edition/reference baseline;A.15.3only for applicable typed planned fillingsA.19.CNCN‑Spec governance cardG.0CG‑Spec legality gateA.19.CHRCHR suite boundary and “governance cards and legality gates are cited as pins, not copied locally” disciplineC.23SoS‑LOG (tri‑state branches; sandbox/probe‑only)F.17UTS (identifier registry; alias/deprecation discipline)F.15static and regression conformance checks for unification slices
-
Used by:
-
Constrains:
- Part‑G authoring: no shadow specs, no silent scalarization, tri‑state guards, penalties routing, typed RSCR causes, defaults with one governing definition, and ID‑continuity refactors.
G.Core:End
G.0 - Define Admissible Comparison and Aggregation for a Frame (CG-Spec)
Status: Stable
Tag. Architectural pattern (foundational Standard; constrains G.1–G.5)
Stage. Design time. Define the admissible comparisons and evidence minima that run-time checks will use.
Primary output. CG‑Spec specifies which comparisons and aggregations a CG‑Frame permits, independently of notation. Publish the specification and its edition pins in UTS for downstream reproducibility and RSCR.
Primary hooks. U.ClaimScope (G), admitted task and object kinds, entityOfConcern, SCP, MinimalEvidence, CNSpecRef, Γ‑fold, Φ(CL) / Φ_plane policy pins, UTS publication (Name Cards + edition pins).
G.0:1 - Problem frame
A team defines or evolves a CG‑Frame (e.g., a frame for creativity measurement, decision quality, architecture trade‑offs, or selected-set publication). Downstream comparison, aggregation, and publication of CHR‑typed observations (G.1–G.5 and beyond) must be:
- lawful with respect to measurement admissibility (scale/unit/polarity constraints),
- auditable with explicit evidence minima and provenance,
- reproducible via pinned editions and explicit policy ids,
- portable under the explicit correspondence and plane-relation conditions that the receiving use requires.
CG‑Spec is the single design-time object that fixes what comparisons and aggregations are lawful in this frame, under which pinned assumptions and minimal evidence requirements, so that run-time selection and publication can be audited without inventing new “local legality gates”.
G.0:2 - Problem
Without a single, frame-level legality standard:
- comparisons and aggregations drift into implicit assumptions (hidden scalarisation; silent totalisation of partial orders),
- numeric gates run on “whatever is available” rather than declared evidence minima and lane/carrier requirements,
- cross-context reuse happens without explicit crossing visibility and stated losses,
- selection outcomes become hard to audit because legality, evidence minima, and penalty routing are not pinned and traceable.
G.0:3 - Forces
- Pluralism vs. comparability. Multiple traditions must co-exist while allowing admissible comparison where justified.
- Expressiveness vs. safety. Rich comparator sets and aggregators vs. measurement admissibility constraints.
- Locality vs. portability. Preserve context-local meanings and establish the correspondences and use conditions needed for transfer, including any justified loss.
- Assurance vs. agility. Evidence minima must meet the claim threshold while staying light enough to adopt.
- Design-time vs. run-time. Keep legality standards and templates design-time; run-time only cites and applies them.
G.0:4 - Solution — CG‑Spec as the design-time legality gate
CG‑Spec is a notation-independent UTS-published object that, for a given CG‑Frame, defines:
- the ComparatorSet (explicit, finite, typed) permitted in this frame,
- the ScaleComplianceProfile (SCP) that constrains lawful operations per characteristic,
- MinimalEvidence requirements per characteristic (lanes, carriers, any applicable freshness windows or crossing conditions, and failure behavior),
- the frame’s penalty and trust folding wiring (by explicit policy ids and edition pins),
- AcceptanceStubs as design-time templates (thresholds remain governed by CAL, not by CG‑Spec),
- optional method-family hooks (e.g., illumination/QD or explore↔exploit guards) as wiring only, with semantics governed by the corresponding patterns.
CG‑Spec constrains downstream gate checks by being referenced and pinned; it is not itself an admissibility mechanism.
G.0:4.1 - G.Core linkage (normative)
Builds on: G.Core (Part-G core invariants; governing-pattern citation)
GCoreLinkageManifest (normative).
Effective obligations/pins/triggers are computed by union expansion of the referenced ids (per G.Core:4.2).
Profiles/sets + explicit deltas; Nil‑elision applies.
CoreConformanceProfileIds :=GCoreConformanceProfileId.PartG.AuthoringBaseGCoreConformanceProfileId.PartG.TriStateGuardGCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted
CorePinSetIds :=GCorePinSetId.PartG.AuthoringMinimalGCorePinSetId.PartG.CrossingVisibilityPins
CorePinsRequired :=(delta over PinSets)UTSRowId[]ReferenceMapComparatorSetRef.editionSCPRef.editionΓFoldRef.edition?MinimalEvidenceRef.edition?FailureBehaviorPolicyId?
DefaultsConsumed := {DefaultId.GammaFoldForR_eff}(governing definition:CC‑G5.4perG.Core.DefaultGoverningDefinitionIndex)RSCRTriggerSetIds := {GCoreTriggerSetId.CGSpecGate}RSCRTriggerKindIds :=(delta over TriggerSets)RSCRTriggerKindId.EvidenceSurfaceEditRSCRTriggerKindId.TokenizationOrNameChangeRSCRTriggerKindId.DefaultGoverningDefinitionChange
TriggerAliasMapRef := ∅
G.0:4.2 - CG‑Spec object model (normative)
CG‑Spec is authored per CG‑Frame. It SHALL:
- be published to UTS as a notation-independent object,
- reference CHR characteristics by id (measurement semantics remain governed by CHR packs),
- constrain what comparisons and aggregations are lawful in this frame via explicit comparator specs and SCP bindings,
- declare minimal evidence gates per characteristic, including explicit failure behavior wiring,
- cite
CN‑Specfor normalization/comparability policies (no duplication and no shadow specs), - publish edition pins and policy ids so downstream selection, parity, shipping, and refresh can be reproducible and RSCR-aware.
G.0:4.3 - CG‑Spec conceptual model (normative)
The Scope entry groups two declarations. The claim scope answers whether the intended comparison applies under the conditions of a particular use. The task and object kinds specify what may enter the comparison. Test the proposed context slice against the former and the proposed task and object against the latter. C.22 calls the claim-scope field in its TaskSignature ScopeSlice(G).
CG‑Spec :=
⟨
UTS.id, Edition,
Context, Purpose, Audience,
Scope := ⟨ClaimScope: U.ClaimScope, Boundary: {TaskKinds, ObjectKinds}⟩,
entityOfConcern, // entity judged, or the identified targets required by the comparison
GroundingHolon?, ReferencePlane ∈ {world|concept|episteme}, // empirical ground when used, and the reference plane
WorldRegime? ∈ {prep|live}, // refines ReferencePlane=world only
ReferenceMap := minimal map{term/id → UTS|CHR|SoTA-pack refs},
CNSpecRef := ⟨CN‑Spec ref, CNSpecRef.edition⟩, // A.19.CN supplies the sole definition of the CN‑Spec governance card
Characteristics := [CHR.Characteristic.id…], // pointers only; authored in G.3 CHR pack
// Edition-addressable segments (pins MUST be exposed)
ComparatorSet := ⟨ComparatorSetId, ComparatorSetRef.edition, [ComparatorSpec…]⟩,
SCP := ⟨SCPId, SCPRef.edition, map Characteristic.id → SCPEntry⟩,
MinimalEvidence := ⟨MinEvId, MinimalEvidenceRef.edition?, map Characteristic.id → MinEvidenceEntry⟩, // min pin: CGSpecRef.edition
Γ‑fold := ⟨GammaFoldId, ΓFoldRef.edition?, // pin the actual model/policy when a numerical fold is used
defaultRef := DefaultId.GammaFoldForR_eff,
override? := ⟨overrideRef, proof_refs, boundary_notes⟩
⟩,
// Penalty routing and plane policies are by explicit policy ids.
// For semantics (tri-state, receiving-use loss, model-bound assurance penalties, crossing visibility, comparison and selection results), cite G.Core and its governing definitions. The policy fields below are conditional; supply every reference required by the selected use or model.
CL‑Routing? := ⟨policy_id, map Bridge.CL → penalty_spec⟩, // when the selected use or model maps Bridge.CL to a penalty
Φ? := ⟨phi_policy_id?, phi_table_ref?, psi_policy_id?, phi_plane_policy_id?⟩, // when any of these policies is used
AcceptanceStubs := [AcceptanceStubId…], // templates only; thresholds remain governed by CAL (G.4)
// These optional hooks specify wiring. The referenced method patterns define their semantics.
E/E‑LOG Guard? := ⟨policy_id, pins…⟩,
Illumination? := ⟨
Q_refs ⊆ Characteristics, D_refs ⊆ Characteristics,
DescriptorMapRef.edition?, DistanceDefRef.edition?, DHCMethodRef.edition?,
InsertionPolicyRef?, PromotionPolicyId?
⟩,
RSCR := ⟨
RSCRTestId[]?, // SHOULD cover: illegal_op_refusals; unit and scale legality checks; freshness windows; // partial-order scalarisation refusals; threshold semantics; declared-model penalty calculations when used (including a selected CL→R_eff model);
// and refusal of degrade.order on unit mismatches (MM‑CHR).
RSCRTriggerKindId[]
⟩,
Naming := UTS Name Cards (designations under F.17/F.18 plus bridge notes),
PublicIdContinuity := ⟨governing definition, DRR link, applicable refresh triggers, refresh cadence?, decay and aging policy?, deprecations⟩,
Provenance := ⟨carrier types, SoTA-pack refs, DRR/SCR linkage⟩
⟩
Local typing notes (normative, non-exhaustive).
ComparatorSpecMUST be typed against SCP/CHR constraints. The frame author chooses the permitted comparators. Examples include dominance where lawful, lexicographic comparison over typed traits, a medoid or median for ordinal data where lawful, and explicit weighted sums only when legality is proven and units are aligned.MinimalEvidenceEntryMUST declare: lane requirements, evidence carriers, freshness window (if any), and explicit failure behavior wiring. The semantics of{pass|degrade|abstain}anddegrade(mode=…)are delegated toG.Core.
G.0:4.4 - Interfaces (normative)
| Interface | Consumes | Produces / constrains |
|---|---|---|
| G.0‑1 Charter | CG‑Frame brief, USM scope signals | CG‑Spec.Scope, entityOfConcern, ReferenceMap |
| G.0‑2 SCP | CHR pack refs (G.3), legality proofs | CG‑Spec.SCP + bindings to lawful operators/aggregators |
| G.0‑3 Evidence | SoTA inputs (G.2), source and carrier provenance account (A.10) | CG‑Spec.MinimalEvidence, Γ‑fold segment pins, and any CL‑Routing or Φ ids required by the selected use or model |
| G.0‑4 Publish | All above | Versioned CG‑Spec@UTS plus Name Cards, public-id continuity records, and RSCR tests and trigger kinds |
| G.0‑5 Expose_CrossingHooks | CG‑Spec + pins required by an applicable E.18 flow crossing or A.21 gate | Inputs for that crossing’s GateChecks: the required plane, intent, lane and lexical SD checks |
| → G.1 | CG‑Spec | Generator guardrails (Comparator/SCP/MinEv pins); degrade/abstain wiring |
| → G.2 | CG‑Spec | Harvesting inclusion/exclusion and crossing policy constraints |
| → G.3 | CG‑Spec | Required CHR characteristics/scales/operators to exist |
| → G.4 | CG‑Spec | Acceptance templates; evidence minima; Γ‑fold override proof hooks |
| → G.5 | CG‑Spec | Eligibility gates and explainability pins (Path/UTS/policy ids) |
| → G.6 | CG‑Spec | EvidenceGraph/SCR pinning surface (policy ids + Path/PathSlice discipline) |
G.0:4.5 - CG‑Spec authoring chassis (informative)
Begin with the receiving question and the available material: the task or decision request, descriptions of the subject and operating conditions, binding requirements, candidate methods, and existing observations. Reuse an adequate frame. When its criteria are not yet known, establish them before filling the specification.
G.0:4.5.1 - Obtain the comparison basis
-
Recover the decision and its conditions. Use C.39.RQ:4.1 to work backward from what the receiving result must enable and forward from the available ways of obtaining it. Inspect the request, relevant commitments or rules, and the descriptions of actual use. For each consequential success or failure, ask which property of the subject makes that difference. Recover a binding limit from its applicable requirement or authorized allocation; an observed value establishes performance under its conditions, not the limit the receiver must accept. Where two uses need different conditions, retain distinct scope slices.
Choose the whole and the model needed for this decision. Start with a decision the receiver could make differently. Keep the properties of the chosen parts fixed and vary their dependencies on other parts or on a larger whole. If the decision changes, include the missing participant or account for its effect at the boundary. Trace the corresponding dependency separately for the system, its method, the performed work or another kind on which the decision relies. Within a fixed whole, ask whether two admissible arrangements look the same in the current model yet call for different actions. Determine what the model must then distinguish.
Stop when the grounds suffice for the decision and range actually promised. A property of each member, a dependency between members and a property of their joint operation have different grounds. Preserve any shared resource, condition or handover that can change the answer. Distinguish the facts and modelling assumptions from the criterion the receiver wants to test. For a bound on actual behaviour, at least one configuration must fit the evidence and assumptions. A bound that settles the decision throughout that range can suffice. See C.37.MM:4.1.1 for choosing and connecting the ranges.
Set the feasibility criterion for the Method at the scale being judged. One permitted way of carrying it out that achieves the required result establishes feasibility. To rule out the Method under the stated conditions, show that every way it admits fails, possibly because all need the same missing contribution. B.1.5:4.1 identifies what the whole Method requires and where it allows substitutes. If every successful execution would need a contribution that none of those ways can supply, success of the whole would contradict its own construction. A jointly workable substitute that delivers the required result settles feasibility positively. Separate successes of the constituents still need compatible handovers, resources and operating conditions to establish a way of carrying out the whole. This reasoning applies at each scale where its premises hold. A plan for two jobs describes proposed work. Their number and duration alone do not identify two constituent Methods. To establish that the required result was obtained, identify the dated execution of the specified Method under the relevant conditions and the evidence for its result. An existing report can already establish both.
Changing the model changes which conclusions carry over. If every success under the concrete conditions would also count as success in the abstract model, ruling out every abstract success rules out every concrete success. To establish concrete feasibility, justify that at least one admissible execution meets the concrete criterion. The abstract model can admit a combination its concrete counterpart rules out. B.5.MR:4.2 tests a claim of model refinement. For actual operating conditions, one genuinely admissible execution that meets the criterion establishes possibility. Impossibility or a guarantee requires grounds covering every relevant way of carrying out the Method. Mutually exclusive established requirements can exclude success in every way. If two defensible readings of a record give different answers, resolve the discrepancy or give the conditional answers. If at least one reading is consistent with the evidence and every admissible reading settles the criterion in the same direction, it can be answered with that input still unknown. A supported negative answer can complete an order to determine feasibility. An order to deliver the result or develop an alternative calls for further work.
-
Turn the differences into characteristics. A.19.ECS:4.2 supplies the construction: start with cases that would satisfy, fail or fall outside the use; identify the properties that distinguish them; split a compound property when its components can vary independently; and establish each characteristic’s subject, meaning, scale, units and observation procedure through G.3/C.16. Existing domain definitions and validated procedures can supply these results directly. If the needed characteristic or procedure is unavailable, return that particular construction to its competent supplier, with the decision it prevents.
-
Choose the comparisons the question needs. For each proposed comparison, state what relation between results would change the receiving decision. Match that relation to the characteristics’ lawful operations: equality of categories, an order where justified, or a numerical operation whose scale, units and assumptions permit it. Retain separate coordinates when trade-offs have not been resolved. A weighted total requires a justified aggregation and the receiving preference basis. G.0 then enumerates this finite set of comparators and its legality conditions.
-
Obtain the evidence minima from the claim. Follow each required characteristic and comparator back to what would establish its inputs and applicability: for example, a traceable observation under the relevant operating conditions, an exact proof for a formal claim, or a justified transfer from a different population. Reuse the applicable domain qualification, measurement or acceptance method. Specify the carriers, conditions and checks it requires, and the consequence when they are missing or fail. G.2 can recover missing source support; G.4 carries the acceptance policy. A currentness window is needed when the claim’s use depends on it; a numerical support fold needs its actual model under B.3/C.2.2.
-
Return the supported frame and its limit. The available material may establish all required criteria, disqualify an operation, or leave a particular criterion, authority, measurement procedure or transfer premise unresolved. Keep those outcomes separate. Identify the source, responsible authority, observation or method construction that could settle an unresolved premise, and which comparison must wait for it. A descriptive comparison can remain useful when a required acceptance judgement is unavailable; it must retain that receiving-use limit.
Worked construction. A team must compare compression methods for a specified archive batch on one host. The available request says “choose a method that fits the backup operation”; the maintenance allocation gives a 15-minute window and a 2 GiB memory allowance, while the archive requirement demands lossless recovery. The batch inventory and host configuration establish the scope. Working backward exposes three required properties: recovery identity, elapsed time and peak memory. Recovery identity is tested by comparing the recovered contents with the original batch; time and memory use the stated measurement procedures and units. The requirements supply the pass limits. Among passing alternatives, an additional request to reduce storage permits comparison by compressed size; without a preference for trading time against size, the frame retains those coordinates separately.
A report that tested recovery on the same batch supplies that observation; a report giving a 12-minute mean on a different host needs a justified transfer or a measurement on the receiving host. That mean does not establish a guarantee of fitting the allocated window. If the receiving requirement is a guaranteed upper bound, the chosen qualification method must supply that bound; otherwise fit remains unestablished. If the allocation is absent, the team can compare observed times and memory but must obtain the applicable allocation before claiming acceptance. If the receiving question changes to archival space efficiency with no execution-window requirement, the performance observations remain usable while the 15-minute gate no longer governs that comparison.
If the same backup requires two nine-minute jobs to finish in fifteen minutes, they must overlap for at least three minutes. Each job peaks at 1.2 GiB. Their peak sum of 2.4 GiB alone leaves fit under the 2 GiB limit open. If each needs 1.2 GiB throughout its run, the necessary overlap cannot fit. If both can use 0.6 GiB during the same three-minute overlap and each fits alone for its remaining six minutes, that schedule fits. This establishes a feasible schedule. A guarantee under other permitted schedules needs grounds covering them.
G.0:4.5.2 - Bind and publish the established basis
- Charter the frame. Declare
Context,Scope,entityOfConcern, boundary examples/non-examples andReferenceMapfrom the receiving question and established conditions. For an earlier kit use G.Core:4.2.3 to recover the subject. A test stand can supply data about a component on it or the method used to inspect that component. - Bind ComparatorSet, SCP and characteristics. Enumerate the permitted comparator forms obtained above. Bind each to the CHR characteristics it uses and their scale, unit, polarity and other legality conditions. Identify each guard bound to a comparator by its reference or pin. Cite existing characteristic ids or publish new ones under UTS discipline.
- Declare MinimalEvidence. For each characteristic, bind the required lanes/carriers, any applicable freshness window or crossing allowance, and explicit failure behavior (tri-state semantics delegated to
G.Core). Keep unresolved prerequisites visible with the operations they prevent. - Pin the support-composition basis. Cite
DefaultId.GammaFoldForR_efffor the model-qualified rule. When using a numerical fold or loss, pin its receiving model, compatible inputs, dependency assumptions and references that justify it; monotonicity and boundedness alone are insufficient. When no common aggregate is justified, the referenced rule retains separate support and a bounded synthesis. Publish any actually used Φ/CL policy ids; keep acceptance thresholds in G.4. - Publish and register regression tests. Publish
CG‑Spec@UTSwith edition-pinned segments; register RSCR tests for the frame’s legality surfaces and evidence minima. Preserve public-id continuity and deprecation notes. State the refresh triggers and any cadence justified by the frame’s use; emit canonical trigger ids when their conditions obtain.
G.0:4.6 - Extensions (pattern-scoped; non-core)
All blocks below are GPatternExtension modules (PatternScopeId; not new PatternIds). They store wiring only and cite governing patterns.
GPatternExtension: SpecRefSurfaces
-
PatternScopeId:
G.0:Ext.SpecRefSurfaces -
GPatternExtensionId:
SpecRefSurfaces -
GPatternExtensionKind:
InteropSpecific -
GoverningPatternId:
A.19.CN -
Uses:
{A.19.CN} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum):
CNSpecRef.edition(and any CN-side policy ids referenced byCG‑Specfields)
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.LegalitySurfaceEdit} -
Notes (wiring-only):
CG‑SpecSHALL cite CN‑Spec; it SHALL NOT restate normalization/comparability semantics.
GPatternExtension: BridgeAndCLWiring
-
PatternScopeId:
G.0:Ext.BridgeAndCLWiring -
GPatternExtensionId:
BridgeAndCLWiring -
GPatternExtensionKind:
InteropSpecific -
GoverningPatternId:
F.9 -
Uses:
{F.9, G.7} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum):
BridgeCardId/BridgeId(when the use relies on the corresponding semantic bridge)CL/CL^kandΦ/Φ_planepolicy ids (when the corresponding calibration or loss model is used)
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.ReferencePlaneEdit} -
Notes (wiring-only): Crossing semantics and penalty routing are delegated to
G.Core; this module only lists the required pins used byCG‑Specentries.
GPatternExtension: SoTAPaletteInputs
-
PatternScopeId:
G.0:Ext.SoTAPaletteInputs -
GPatternExtensionId:
SoTAPaletteInputs -
GPatternExtensionKind:
DisciplineSpecific -
GoverningPatternId:
G.2 -
Uses:
{G.2} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum):
SoTA-Pack@CG‑Framerefs used to justify comparator admissibility, evidence minima, and crossing allowances (e.g., claim sheets, operator inventory, bridge matrix ids)
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.FreshnessOrDecayEvent} -
Notes (wiring-only): Any SoTA palette/tradition semantics are governed by
G.2.G.0only requires thatCG‑Specentries cite the needed SoTA artefacts for auditability.
GPatternExtension: QDAndExplorationHooks
-
PatternScopeId:
G.0:Ext.QDAndExplorationHooks -
GPatternExtensionId:
QDAndExplorationHooks -
GPatternExtensionKind:
MethodSpecific -
GoverningPatternId:
C.18 -
Uses:
{C.18, C.19, C.23} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum):
DescriptorMapRef.edition?,DistanceDefRef.edition?,InsertionPolicyRef?FailureBehaviorPolicyId/ SoS‑LOG branch policy id whendegrade(mode=…)is used
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent} -
Notes (wiring-only):
CG‑Specmay declare QD/exploration hooks. The referenced method patterns define their semantics.
G.0:5 - Archetypal Grounding — Tell–Show–Show; System / Episteme
G.0:5.1 - Archetype 1: System comparability under mixed evidence and unit constraints
Tell. Two labs compare energy efficiency results of a physical system where measurements use different rigs and units, and some evidence is missing.
Show (failure without CG‑Spec). The team averages an ordinal safety rating, mixes units (“kWh” vs “MJ”), and silently treats missing lanes as zeros. Cross-lab reuse also leaves any differences in measurement meaning and their effects unexplained.
Show (repair with CG‑Spec). A conformant CG‑Spec:
- pins the lawful comparator(s) (e.g., unit-aligned ratio comparisons only; ordinal comparisons are order-only),
- declares the
MinimalEvidencelanes and carriers for each characteristic, together with the freshness windows used in this lab comparison, - declares explicit failure behavior wiring (tri-state semantics delegated to
G.Core), - exposes Bridge and bounded-use pins when reuse projects between different local meanings under A.19.CN/F.9; unit conversion within the declared comparison basis follows CN-Spec and SCP,
- publishes the pinned editions so parity/refresh can detect drift.
G.0:5.2 - Archetype 2: Epistemic comparability for selected-set publication across traditions
Tell. A team selects an R&D set using multiple evaluation traditions: safety assurance, cost models, and readiness heuristics.
Show (failure without CG‑Spec). The team collapses partial orders into a single score, hides the threshold policy in code, and cannot explain why cross-tradition penalties changed between runs.
Show (repair with CG‑Spec). A conformant CG‑Spec:
- defines a comparator bundle (e.g., Pareto dominance + explicit lexicographic tiebreaks where lawful),
- pins
CNSpecRef.editionand the editioned segments (ComparatorSetRef.edition,SCPRef.edition,MinimalEvidenceRef.edition), - makes
AcceptanceStubsexplicit as templates while locating thresholds in CAL (G.4), - ensures RSCR triggers are emitted when comparator or policy pins change.
G.0:6 - Bias-Annotation
CG‑Spec can encode (and therefore amplify) biases if authored carelessly:
- Tradition favoritism. Comparator choices may privilege a tradition’s evidence style; mitigation: declare evidence minima and any justified crossing loss, and require grounds for cross-tradition aggregation.
- Metric gaming and Goodhart effects. A single score can conceal trade-offs. Preserve the comparator’s declared relations and the selector’s actual outcome. State the authority and basis for a scalar ranking used in a decision.
- Hidden thresholds and opaque safety policy. Embedding acceptance thresholds in prose or code hides value judgments; mitigation: keep thresholds in CAL acceptance clauses and pin policy ids.
- Scope creep. A comparison may be reused for a different entityOfConcern or reference plane. Retain
entityOfConcernandReferencePlanepins, and apply the governing rule for any plane relation the comparison uses. Include E.18/A.21 crossing or gate pins when that use applies.
G.0:7 - Conformance Checklist (normative)
| ConformanceId | Statement |
|---|---|
| CC‑G0‑CoreRef | G.0 is conformant only if the applicable core obligations listed in G.0:4.1 are satisfied (delegation to CC‑GCORE‑*; no shadow specs, no competing defaults, typed RSCR triggers, explicit pins). |
| CC‑G0‑01 | CG‑Spec is published as a notation-independent UTS object with explicit Edition, Context, Scope, entityOfConcern, and a minimum ReferenceMap. |
| CC‑G0‑02 | CNSpecRef.edition is present and identifies the edition of the external governance card cited by CNSpecRef (no local redefinition of CN semantics). (Delegation target: CC‑GCORE‑CN‑CG‑1.) |
| CC‑G0‑03 | ComparatorSet is explicit and finite; each comparator is typed and bound to SCP and referenced CHR characteristics; anything not enumerated MUST be treated as illegal/abstain by default (no implicit comparator defaults). |
| CC‑G0‑04 | SCP declares, per characteristic, the lawful operation regime needed for each referenced comparator (scale/unit/polarity constraints and any required proofs/refs). |
| CC‑G0‑05 | MinimalEvidence is declared per characteristic and includes explicit lane/carrier requirements, freshness window references (if any), and explicit failure behavior wiring (tri-state semantics delegated). If freshness windows are used, a stable reference making the applicable window recoverable (e.g., a PathSliceId with its declared time window) MUST be pinned for audit. |
| CC‑G0‑06 | The edition-addressable Γ-fold segment MUST cite DefaultId.GammaFoldForR_eff at G.5 CC‑G5.4. Pin any numerical model/policy actually used, with its receiving quantity, input meanings and scales, dependencies, operation and proof/justification refs as required by that governing rule and B.3/C.2.2. With no justified numerical fold, retain the separate support permitted by the governing rule; the segment is not a demand for a score. |
| CC‑G0‑07 | If crossing penalties are used, all policy ids used by the selected model are explicit and auditable (exposed as pins/refs, including applicable CL‑Routing and Φ entries) and are required pins for downstream SCR publication on penalised claims (see G.6). |
| CC‑G0‑08 | AcceptanceStubs in CG‑Spec are templates only; any context-local thresholds/acceptance policies are governed by CAL acceptance artefacts (G.4) and are cited, not duplicated. |
| CC‑G0‑09 | RSCR tests and triggers for edits to legality surfaces and evidence minima are present and use canonical RSCRTriggerKindIds. The RSCR test set SHOULD cover at least: illegal_op_refusals; unit and scale legality checks; freshness windows; partial-order scalarisation refusals; threshold semantics; declared-model penalty calculations when used (including a selected CL→R_eff model); refusal of degrade.order on unit mismatches (MM‑CHR). |
| CC‑G0‑10 | PublicIdContinuity identifies the governing definition, DRR link and deprecations. State applicable refresh triggers and any use-justified cadence or aging policy. Deprecations preserve lexical continuity (Δ-discipline; delegated to CC‑GCORE‑ID‑*). |
| CC‑G0‑11 | (Conditional) If Illumination / QD hooks are present, DescriptorMapRef.edition, DistanceDefRef.edition, and any InsertionPolicyRef / promotion policy ids are pinned (or explicitly marked absent) and are recorded in provenance/audit pins. |
| CC‑G0‑12 | (Conditional) If freshness windows influence gating/selection, they are published and enforced, and references making the relevant windows recoverable (PathSliceId with a declared time window, or equivalent) are recorded in SCR/audit pins. |
| CC‑G0‑13 | Pre-flight numeric gates. Any numeric comparison/aggregation declared in ComparatorSet has associated GateChecks for unit legality, scale legality, pinned SOP/editions, and declared comparability assumptions; failing any check requires refusing the operation or returning an abstain guard result (tri-state semantics delegated). |
| CC‑G0‑14 | GateCrossing hook exposure. When an export is consumed by an independently applicable E.18 flow crossing or A.21 gate, provide its required Expose_CrossingHooks inputs. Failure of an applicable check blocks that crossing or the publication that requires it. |
| CC‑G0‑Φ | An actually used numerical Φ(CL) or plane-loss policy has a justified receiving quantity, scale, input interpretation, assumptions, and derivation or calibration under B.3/C.2.2. Publish the policy ids and any required table. Preserve the model’s loss direction and bounds; monotonicity, boundedness, or clipping alone does not establish a valid loss model. |
| CC‑G0‑Unknowns | Delegated. Unknown handling MUST follow the tri-state guard semantics `{pass |
| CC‑G0‑CSLC | Scale/unit/polarity legality MUST be proven before any aggregation; illegal arithmetic on ordinal/nominal values is nonconformant. (Governed by the relevant legality patterns; G.0 only binds and cites.) |
G.0:8 - Common Anti-Patterns and How to Avoid Them
- Anti-pattern: shadow legality gates in downstream code. Avoid by requiring downstream to cite
CG‑Specsegments by id+edition. - Anti-pattern: “one number to rule them all”. Report the pairwise relation when the order asks about two candidates. Identify a frontier under the dominance rule for its candidate domain. For a selection apply the declared selection rule. Combining several criteria into a scalar requires an explicit, lawful operation. Comparing a single measured size can use its own scale directly.
- Anti-pattern: thresholds inside CG‑Spec or CHR. Avoid by keeping thresholds and acceptance logic in CAL and citing from
CG‑Speconly via stubs/templates. - Anti-pattern: unqualified cross-context reuse. Expose the correspondence, calibration, loss-policy and reference-plane pins required by the receiving use (
G.Core:4.2.3).
G.0:9 - Consequences
- Lawful comparability. The frame enumerates the comparisons and aggregations it permits and the constraints on each.
- Auditable selection. Downstream selectors can justify outcomes via pinned legality surfaces and explicit evidence minima.
- Explicit transfer conditions. Cross-context reuse exposes the correspondence, receiving conditions and any justified loss on which it relies.
- Lower drift under evolution. Compare edition pins to detect a change in the comparison basis. Use the corresponding typed RSCR trigger to refresh the affected comparisons.
G.0:10 - Rationale
CG‑Spec centralises frame-level comparability constraints so that:
- CHR authorship (G.3) remains about measurement meaning rather than implicit thresholds,
- CAL (G.4) governs context-local acceptance/threshold policies and proof ledgers,
- selectors and dispatchers (G.5) remain policy-governed and auditable rather than encoding hidden legality assumptions,
- refresh (G.11) can treat legality edits and pin changes as explicit causes with canonical trigger ids.
G.0:11 - SoTA‑Echoing
The pattern connects the following evaluation and governance practices:
- treating abstention or deferral as an evaluation outcome, for example when the available evidence does not warrant the requested judgement (as in selective prediction),
- preserving the declared result of a multiobjective comparison, including established incomparability, and a frontier or archive when the receiving operation calls for one,
- versioning and pinning the evaluation surfaces and publishing the policy identifiers so the evaluation can be reproduced,
- making evidence minima explicit and auditable (a conceptual analogue of modern reproducibility/robustness checklists and evaluation protocols),
- keeping method-family specifics modular (e.g., QD/archives, open-ended exploration budgets) via explicit wiring to governing patterns rather than embedding method semantics into the universal legality gate.
G.0:12 - Relations
Builds on: G.Core, A.19.CN (CN‑Spec), A.10 (evidence-provenance account), A.17–A.19 / C.16 (MM‑CHR legality), A.18 (CSLC), B.3 (trust / Γ‑fold family), F.* (contexts, bridges, CL, UTS), E.10 (lexical rules), E.5.* (notation independence discipline).
Used by: G.1 (generator guards), G.2 (harvesting constraints), G.3 (required CHR), G.4 (acceptance templates / proof hooks), and G.5 (eligibility gates); G.6 when the represented claim depends on CG-Spec, and parity/shipping/refresh where that specification is pinned.
Publishes to: UTS (Name Cards + editioned CG‑Spec segments).
G.0:End
G.1 - Author a Reusable CG-Frame Generator and Selector Kit
Status: Stable
Tag. architectural pattern; generator chassis (design‑time kit / authoring scaffold)
Normativity. normative, except sections explicitly marked informative
Stage. design‑time authoring of a generator‑kit with a run‑time execution façade (policy‑governed; edition‑aware)
Primary output. Publish the six cards M1…M6 as a reusable CG‑Frame kit. The versioned kit manifest, identified by CGKitId, binds all six by their card ids.
Primary hooks. see §12 Relations (notably G.Core, G.0, G.2, G.5, G.10, G.11)
Working‑model first (informative). prefer working models and didactic micro‑examples; escalate to formal harnesses only when risk warrants (per E.8).
Start here when. You are authoring a reusable generator, selector, or set-result scaffold rather than a one-off plan, one-off comparison, or tool-specific method recipe.
Neighboring FPF patterns. Use G.2 for the local SoTA set, G.5 for governed set-return selection, G.10 for shipping surfaces, G.11 for refresh wiring, and F.17 when the result must also land on a human-facing UTS surface.
Use another pattern when. For a one-off governed comparison, use A.19.CPM to compare an admitted profile pair under an explicit comparator, with G.0 for comparison legality; use G.5 for a shortlist. For project alignment, use A.15. When authoring a kit, make the chassis and its bindings explicit before choosing implementation tooling.
G.1:1 - Problem frame
You are authoring a CG‑Frame and want a repeatable scaffold that connects:
- a declared scope anchor (
CG‑FrameContext,entityOfConcern, governing spec refs), - a local SoTA set (scoped and provenance‑anchored),
- a variant pool (candidate ideas / decision options / method variants),
- a shortlist (a set-result outcome, not a forced singleton),
- publication‑ready bindings into Part‑F artefacts (UTS rows, Name Cards, RSCR tests, worked examples),
- and refresh readiness (telemetry hooks + RSCR wiring).
The chassis connects the cards. The following patterns define their methods and payloads.
- harvesting semantics are governed by
G.2, - selection/dispatch semantics are governed by
G.5, - CHR/CAL payload semantics are governed by
G.3/G.4, - shipping semantics are governed by
G.10, G.11governs refresh orchestration.
G.1:2 - Problem
Without the connections supplied by the chassis, a CG-Frame kit can have the following gaps:
- Unscoped SoTA inputs. A reader cannot reconstruct which sources belong to the local set.
- Generation is ad‑hoc: variant candidates are emitted without a stable trace of why/when/how.
- Selection is opaque: eligibility/acceptance and assurance are not pinned to explicit surfaces.
- Outputs are hard to reuse. The kit lacks the bindings needed to hand them on to UTS, RoleDescription, Concept‑Sets or RSCR.
- No kit‑level snapshot: the scaffold lacks a versioned manifest, so downstream can’t reliably cite “which chassis edition” was used.
- Refresh is unplanned: there is no canonical wiring from edits/telemetry/decay to RSCR causes along the P2W path.
G.1:3 - Forces
- Breadth vs. precision: harvest wide enough to avoid local dogma, but keep the artefact actionable.
- Generativity vs. assurance: encourage novelty while keeping evidence, legality, and trust inspectable.
- Local meaning vs. portability: keep meaning local by default; crossing must be explicit and auditable.
- Expressiveness vs. parsimony: resist inventing new types/slots; prefer reuse and explicit wiring.
- Stability vs. evolution: keep stable IDs and pins while allowing SoTA, policies, and editions to evolve.
- Didactic clarity vs. normative minimalism: authors need a concrete scaffold, but universal invariants must not be duplicated outside
G.Core.
G.1:4 - Solution
G.1:4.1 - G.Core linkage (normative)
// Canonical form: see G.Core (Nil‑elision + Expansion rule for profiles/sets/pin‑sets).
GCoreLinkageManifest := ⟨
CoreConformanceProfileIds := {
GCoreConformanceProfileId.PartG.AuthoringBase,
GCoreConformanceProfileId.PartG.TriStateGuard,
GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted,
GCoreConformanceProfileId.PartG.ShippingBoundary
},
CorePinSetIds := {
GCorePinSetId.PartG.AuthoringMinimal,
GCorePinSetId.PartG.CrossingVisibilityPins
},
// Prefer sets; use deltas for pattern‑specific additions.
RSCRTriggerSetIds := { GCoreTriggerSetId.SoTAHarvestSynthesis },
RSCRTriggerKindIds := { RSCRTriggerKindId.BaselineBindingEdit },
// Kit identifiers governed by this pattern (the “six cards”).
CorePinsRequired := {
SoTAPaletteDescriptionId,
SoTA_SetId,
VariantPoolId,
ShortlistId,
CGFrameLibraryId,
RefreshReadinessCardId,
CGKitId,
// Local pointer-map surface for vocabulary + observables-to-CHR anchoring.
// (May cite `G.0:CG‑Spec.ReferenceMap`; do not duplicate semantics.)
ReferenceMap,
// RSCR regression tests used by the chassis (if any).
RSCRTestId[]?,
// For a planned baseline, identify the A.15.2 WorkPlan and local locator; add declaration ref, member designator and plan-local filling-row locator when A.15.3 applies.
WorkPlanRef[]?
},
// Consumed defaults (each default cites the governing definition listed in `G.Core.DefaultGoverningDefinitionIndex`).
DefaultsConsumed := {
DefaultId.GammaFoldForR_eff, // governing definition: CC-G5.4
DefaultId.PortfolioMode, // governing definition: CC-G5.23
DefaultId.DominanceRegime // governing definition: CC-G5.28
}
⟩
Citation rule (normative): CC‑GCORE‑*, RSCRTriggerKindId.*, and DefaultId.* semantics are governed by their canonical definitions: primarily G.Core, and for the defaults above the definitions listed in G.Core.DefaultGoverningDefinitionIndex. G.1 MUST NOT restate or redefine those semantics.
G.1:4.2 - Six‑module generator chassis (normative)
Core artefact: CGFrameReadyGeneratorKit := ⟨M1, M2, M3, M4, M5, M6⟩, where each Mi is a card with an explicit I/O surface and stable identifiers.
CGKitId identifies the versioned kit manifest (CG‑Kit@CG‑Frame) that lists the six card ids and the minimal wiring pins needed to treat the chassis as a reusable unit (this is not a shipping pack; shipping remains governed by G.10).
The chassis inventories the cards and their connections so readers can view them together. The specifications it cites define CN, CG, CHR, CAL and selection semantics.
G.1:4.2.1 - M1 — CG‑FrameContext Card (scope anchor)
Governs (kit surface):
-
CG‑FrameContextand its binding pins:entityOfConcern, the entity or targets judged, with theGroundingHolonandReferencePlanerequired by the frame (PartG.AuthoringMinimal); a test site can ground claims about different targets.CNSpecRef.edition,CGSpecRef.edition(pin set:PartG.AuthoringMinimal)ReferenceMap(citeG.0:CG‑Spec.ReferenceMap; do not duplicate semantics)- any declared crossing/policy pins (pin set:
PartG.CrossingVisibilityPins)
Purpose: give every downstream card the same scope anchor. If the receiving criteria or comparison basis are not yet established, obtain them through G.0:4.5.1 before binding M1; reuse an adequate existing frame.
Notes: any spec-legality content is cited via A.19.CN (CN‑Spec) and G.0 (CG‑Spec) (delegation target: CC‑GCORE‑CN‑CG‑1 via CC‑G1‑CoreRef); this card does not introduce a local “mini‑spec”.
G.1:4.2.2 - M2 — SoTA_Set@CG‑Frame (harvester output card)
Governs (kit surface):
SoTAPaletteDescriptionIdandSoTA_SetIdbound toCG‑FrameContext- explicit provenance anchors for the set (via
A.10), and any published UTS stubs/rows when applicable
Governing pattern. G.2 defines the harvesting discipline and the SoTA-pack payload.
M2 binds the harvested set to the chassis. When a use relies on coverage, cite G.2’s CoverageJudgementRef with its HarvestPolicy basis, counted units and receiving question. M2 supplies no alternative family count.
G.1:4.2.3 - M3 — VariantPool (candidate inventory + emitter trace)
Governs (kit surface):
VariantPoolIdbound toCG‑FrameContext- per‑candidate minimal traceability fields (emitter identity,
EmitterPolicyRef(policy‑id/ref; defined by the governing pattern), method/generator refs when declared, edition pins, provenance anchors) - optional, per‑candidate assurance preview pointers (e.g.,
PathSliceId?and/orSCRId?when early assurance is recorded) and optional QD/Open‑Ended scaffolding stubs (only when introduced by explicitGPatternExtensionblocks)
Guardrails (via G.Core):
- tri‑state eligibility handling, penalties routing, crossing visibility, and set‑return constraints are not defined here; they are enforced via
G.Coreconformance.
Governing pattern for method payload: method‑specific emitter semantics remain in their governing definitions, cited through Extensions (e.g., the relevant C.17, C.18, and C.19 definitions).
The M3 core MUST remain a method‑agnostic inventory.
G.1:4.2.4 - M4 — Shortlist (selector output)
Governs (kit surface):
ShortlistIdbound toCG‑FrameContext- a selected set of candidates plus rationale and SCR-addressable audit references required by G.5 (
SCRIdrequired;DRRIdoptional; citePathId/PathSliceIdwhen applicable). Add assurance records only when an actual named assurance claim is current. - optional front metadata or archive metadata needed for reproducibility when used: ε‑front parameters and/or archive snapshot hooks, with governing-definition assignment through
G.5/C.18/C.19(no local semantics inG.1)
Governing pattern: selection/dispatch semantics are governed by G.5.
M4 MUST preserve set‑return semantics (as governed by G.Core) and MUST NOT hard‑code a forced singleton outcome.
G.1:4.2.5 - M5 — CG‑FrameLibrary (published bindings index)
Governs (kit surface):
-
CGFrameLibraryIdbound toCG‑FrameContext -
an index of referenced CG‑Frame artefacts ready for reuse:
- CHR/CAL/LOG bundles (by their ids; semantics governed by
G.3,G.4,G.8) - published identifiers (UTS rows, Name Cards) per Part‑F governing definitions
- additional Part‑F binding surfaces (e.g., RoleDescription templates, Concept‑Set rows) by ids locating those surfaces under their governing definitions
- RSCR test identifiers (e.g., from
F.15) and worked examples (where applicable)
- CHR/CAL/LOG bundles (by their ids; semantics governed by
Boundary: M5 is a kit/library surface, not shipping. If a shipped pack is needed, governing-definition assignment is G.10.
G.1:4.2.6 - M6 — RefreshReadiness Card (telemetry hooks + wiring)
Governs (kit surface):
RefreshReadinessCardIdbound toCGFrameLibraryId(and thus toCG‑FrameContext)CGKitIdidentifies the versioned kit manifest that bindsM1…M6into a single reusable unit. The manifest MUST enumerate the card ids and MAY cite deprecations or edition changes issued under the governing definitions.- declared telemetry hooks (what signals are observed, with what pins)
- declared RSCR wiring: which
RSCRTriggerKindIdare relevant (canonical ids), with minimal required payload pins. For a planned baseline, include theA.15.2WorkPlan reference and its local baseline locator; filling an independently declared position underA.15.3also requires the governing declaration reference, member designator and plan-local filling-row locator.
Boundary: orchestration semantics are governed by G.11.
M6 prepares refresh‑readiness metadata and wiring stubs; it does not define scheduling/priority heuristics.
G.1:4.3 - Minimal I/O surface (normative)
| Module | Consumes | Produces |
|---|---|---|
| M1 | CG‑Frame brief + entityOfConcern + CNSpecRef/CGSpecRef (edition‑pinned) | CG‑FrameContext + context pins |
| M2 | discovery inputs + inclusion criteria (via G.2) | SoTA_SetId (+ provenance anchors; optional UTS stubs/rows) |
| M3 | SoTA_SetId + local constraints + emitter policy pins (via Extensions) | VariantPoolId (+ candidate trace/provenance; optional method payload via Extensions) |
| M4 | VariantPoolId + acceptance/eligibility surfaces (via G.4/G.5) | ShortlistId (selected set / set-result) + rationale refs |
| M5 | ShortlistId + CHR/CAL/LOG bundle refs + UTS/Name refs | CGFrameLibraryId (library index; publish‑ready bindings) |
| M6 | telemetry inputs + applicable freshness/decay policy pins + RSCR tests | CGKitId + RefreshReadinessCardId (wiring to G.11; no orchestration governance) |
G.1:4.4 - Extensions (pattern‑scoped; non‑core)
All method/discipline/generator specifics MUST be expressed as GPatternExtension blocks.
Guard:
G.1:Ext.*are PatternScopeId values (internal, pattern‑scoped), not new patterns and not newPatternId.
G.1:4.4.1 - GPatternExtension — G.1:Ext.HarvesterWiring
PatternScopeId: G.1:Ext.HarvesterWiring
GPatternExtensionId: HarvesterWiring
GPatternExtensionKind: GeneratorSpecific
GoverningPatternId: G.2
Uses: {G.2}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
SoTAPaletteDescriptionIdSoTA_SetIdCoverageJudgementRefand itsHarvestPolicyRefwhen coverage is consumedClaimSheetId[]/BridgeMatrixId(as referenced by the chosen G.2 pack form)CNSpecRef.edition,CGSpecRef.edition(already required viaGCorePinSetId.PartG.AuthoringMinimal)
RSCRTriggerSetIds: {GCoreTriggerSetId.SoTAHarvestSynthesis}
Notes (wiring‑only). G.2 defines harvesting semantics, including living-review funnels, inclusion policy families and SoS indicator families.
G.1:4.4.2 - GPatternExtension — G.1:Ext.ShortlistWiring
PatternScopeId: G.1:Ext.ShortlistWiring
GPatternExtensionId: ShortlistWiring
GPatternExtensionKind: MethodSpecific
GoverningPatternId: G.5
Uses: {G.5, G.4}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
ShortlistIdSCRId(selector audit reference under G.5; its record carries assurance only when an actual named assurance claim is current under B.3)DRRId?(only when a decision‑rationale artefact is minted)TaskSignatureRef?(only for task‑templated selection)AcceptanceClauseId[](as referenced fromG.4outputs)- any explicit selector policy pins (policy‑id/ref; defined by the governing pattern) when not defaulted (the omitted default cites its governing definition through
G.Core.DefaultGoverningDefinitionIndex)
Notes (wiring‑only). M4 binds its output to the G.5 selector/dispatcher kernel.
G.1:4.4.3 - GPatternExtension — G.1:Ext.CreativityCHR
PatternScopeId: G.1:Ext.CreativityCHR
GPatternExtensionId: CreativityCHR
GPatternExtensionKind: DisciplineSpecific
GoverningPatternId: C.17
Uses: {C.17, G.3}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
CHRPackId?(if creativity characteristics are published/typed)- edition/policy pins required by the chosen creativity characteristic set (governed by
C.17)
Notes (wiring‑only): G.1 only records which creativity characteristics are used for M3/M4 wiring; legality/typing lives in the CHR governing definitions.
G.1:4.4.4 - GPatternExtension — G.1:Ext.NQD
PatternScopeId: G.1:Ext.NQD
GPatternExtensionId: NQD
GPatternExtensionKind: MethodSpecific
GoverningPatternId: C.18
Uses: {C.18, C.19}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
DescriptorMapRef.editionDistanceDefRef.editionInsertionPolicyRef(policy id / ref, as defined by the governing definition)TaskSignatureRef?(when QD is enabled via TaskSignature flags/traits rather than by an external switch)DHCMethodRef.edition?(when illumination/coverage summaries are pinned to a method)EmitterPolicyRef(policy‑id/ref; identifies the chosen emitter policy under its governing definition, e.g.,C.19when E/E‑LOG is used)
RSCRTriggerKindIds: {RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent}
Notes (wiring‑only): QD/QD‑adjacent algorithm families and their parameterisations belong to C.18 and C.19; G.1 only fixes the pins needed to make the VariantPool and Shortlist reproducible.
G.1:4.4.5 - GPatternExtension — G.1:Ext.OpenEndedFamilyWiring
PatternScopeId: G.1:Ext.OpenEndedFamilyWiring
GPatternExtensionId: OpenEndedFamilyWiring
GPatternExtensionKind: GeneratorSpecific
GoverningPatternId: G.2 (family semantics are governed by SoTA cards; this block only wires pins; selector-side wiring is governed by G.5.)
Uses: {G.2, G.5, C.19, C.23}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
GeneratorFamilyId[]TransferRulesRef.edition(mandatory when Open‑Ended is enabled)EnvironmentValidityRegionRef?CoEvoCouplerRef[]?SoSLogBranchId[]?(when validity of generated tasks is gated by explicit branches)
RSCRTriggerKindIds: {RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent}
Notes (wiring‑only). The extension enables declared sets of {Environment, MethodFamily} pairs. It should cite/align with the selector-side wiring in G.5:Ext.OpenEndedFamilyWiring.
G.1:4.4.6 - GPatternExtension — G.1:Ext.RefreshWiring
PatternScopeId: G.1:Ext.RefreshWiring
GPatternExtensionId: RefreshWiring
GPatternExtensionKind: GeneratorSpecific
GoverningPatternId: G.11
Uses: {G.11}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
RefreshReadinessCardIdRSCRTestId[]- canonical
RSCRTriggerKindId[]emitted/recorded (aliases only as labels, if any)
RSCRTriggerSetIds: {GCoreTriggerSetId.RefreshOrchestration}
Notes (wiring‑only): M6 declares readiness and wiring; orchestration semantics (queueing, prioritisation, cadence) are governed by G.11.
G.1:4.4.7 - GPatternExtension — G.1:Ext.ShippingWiring
PatternScopeId: G.1:Ext.ShippingWiring
GPatternExtensionId: ShippingWiring
GPatternExtensionKind: GeneratorSpecific
GoverningPatternId: G.10
Uses: {G.10}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
CGFrameLibraryIdSoTAPaletteDescriptionId,SoTA_SetIdCHRPackId?,CALPackId?,SoS‑LOGBundleId?,ParityReportId?(as present in the library index)EvidenceGraphId?,BridgeMatrixId?,BridgeCalibrationTableId?(when cited by the shipped artefacts)UTSRowId[]?(when any public ids are minted/published)WorkPlanRef[]?with local baseline locators (when a planned baseline is cited by the shipment surface underA.15.2); include the governing declaration reference, member designator and plan-local filling-row locator whenA.15.3applies
Notes (wiring‑only). When shipping any of the kit’s artefacts, record the minimum wiring from the chassis/library index to G.10 for that shipment.
G.1:5 - Archetypal Grounding — Tell–Show–Show (informative)
Tell. Use the six cards to reconstruct the frame, sources and candidate pool used for a selected set, the artefacts indexed for publication and the changes that can trigger refresh. The manifest identifies the cards in that kit edition.
Show A (R&D multi‑criteria decisions).
- M1: define
CG‑FrameContextfor “R&D decision options”, pinCNSpecRef/CGSpecRefeditions, and publishentityOfConcern+ReferencePlane. - M2: build
SoTA_SetIdviaG.2using a living-review funnel with a declared update cadence and source-selection reporting (for example, PRISMA-style reporting) and publish UTS stubs for reusable constructs. - M3: emit a
VariantPoolIdwhere each candidate cites its emitter policy and provenance; if QD is used, wireDescriptorMapRef.editionandDistanceDefRef.editionviaG.1:Ext.NQD. - M4: produce
ShortlistIdas a selected-set / shortlist surface viaG.5, with acceptance predicates sourced fromG.4. - M5: publish a
CGFrameLibraryIdindexing the chosen CHR/CAL/LOG bundles and UTS rows; register RSCR tests. - M6: declare refresh readiness (telemetry pins + canonical RSCR trigger kinds) and wire to
G.11.
Show B (clinical operations; safety‑first acceptability).
- M1: scope a CG‑Frame around dose adjustment decisions; pin legality and evidence minima explicitly.
- M2: harvest SoTA models and safety constraints as a reconstructible set (governed by
G.2). - M3: generate policy‑constrained candidate protocols; emitter trace and evidence pins are mandatory.
- M4: return the shortlisted protocols as a set, which may have one member. Choosing one from a shortlist of several requires an explicit decision policy.
- M5/M6: publish and wire refresh (decay events, policy changes, and evidence updates retrigger along the P2W path).
G.1:6 - Bias‑Annotation (informative)
- Recency bias. To counter “newest paper wins”, apply the stated inclusion criteria under
G.2and follow any update cadence the review requires. - Novelty bias: over‑rewarding novelty at the expense of legality/assurance (mitigate by making acceptance and assurance pins explicit and governed).
- Algorithmic favoritism: baking a preferred generator into “the chassis” (mitigate by keeping M3 method‑agnostic and putting method‑specific wiring into Extensions).
- Scalarisation bias: collapsing a partial order or selected set into a single score. Follow G.Core for the comparator’s result and retain the actual selector outcome when selection is commissioned.
- Hidden‑crossing bias: unqualified reuse across contexts (mitigate by retaining the correspondence and receiving-use pins required by
G.Core:4.2.3).
G.1:7 - Conformance Checklist (normative)
| ConformanceId | Statement |
|---|---|
| CC‑G1‑CoreRef | The pattern MUST satisfy the effective CoreConformanceIds implied by G.1:4.1 (GCoreConformanceProfileId expansion + deltas), per G.Core expansion rules. |
| CC‑G1‑01 | The reusable CG-Frame kit MUST include all six cards M1…M6 with stable ids and a versioned kit manifest identified by CGKitId, including at minimum {CGKitId, CG‑FrameContext, SoTAPaletteDescriptionId, SoTA_SetId, VariantPoolId, ShortlistId, CGFrameLibraryId, RefreshReadinessCardId}. |
| CC‑G1‑02 | M1 MUST bind the kit to a single CG‑FrameContext and MUST expose the required pins from GCorePinSetId.PartG.AuthoringMinimal (including entityOfConcern and CNSpecRef/CGSpecRef editions). M1 MUST also expose (or explicitly cite) a ReferenceMap surface and MUST NOT restate its semantics (cite G.0:CG‑Spec.ReferenceMap). |
| CC‑G1‑03 | M2 MUST be wired to G.2 (or explicitly cite the G.2 artefacts governed by cited patterns) and MUST be reconstructible as a scoped set, including SoTAPaletteDescriptionId + SoTA_SetId (not free‑floating prose). Provenance MUST be anchored via A.10 for the emitted set. |
| CC‑G1‑04 | M3 MUST record emitter provenance as a wiring surface, including EmitterPolicyRef (policy‑id/ref), edition pins, and provenance anchors (via A.10). Any method‑specific fields MUST be introduced only via GPatternExtension blocks. |
| CC‑G1‑05 | M4 MUST be wired to G.5 (or explicitly cite G.5 artefacts governed by cited patterns) and MUST preserve set-result outcomes. SCRId MUST be present (or recoverable from an explicitly cited SCR record) so the G.5 audit references are addressable; assurance content is required only for an actual named assurance claim; DRRId SHOULD be present when a decision‑rationale artefact is minted. |
| CC‑G1‑06 | M5 MUST publish a library/index surface that points to referenced CHR/CAL/LOG artefacts and to any minted public ids (UTSRowId[], Name Cards) via the canonical governing definitions (Part F), without introducing shadow specs (delegation target: CC‑GCORE‑CN‑CG‑1 via CC‑G1‑CoreRef). |
| CC‑G1‑07 | M6 MUST publish CGKitId and expose refresh-readiness wiring: canonical RSCRTriggerKindId[] applicability, minimal payload pins and RSCR test ids. Planned baseline references MUST follow G.1:4.2.6, including the additional references required when A.15.3 applies; orchestration semantics MUST be cited to G.11. |
| CC‑G1‑08 | Any method/discipline/generator specificity in G.1 MUST be located in G.1:4.4 as GPatternExtension blocks with PatternScopeId, GPatternExtensionKind, and GoverningPatternId. If QD/illumination or Open‑Ended generator families are declared, the corresponding extension blocks MUST be present and MUST carry the edition and policy pins required by the governing pattern. |
G.1:8 - Common Anti‑Patterns and How to Avoid Them (informative)
-
Anti‑pattern: “Shadow CN/CG spec inside the chassis.” Avoid: keep CN/CG as cited governing spec refs; use pins and governing definition references only.
-
Anti‑pattern: “Chassis hard‑codes a favourite algorithm.” Avoid: keep M3 core method‑agnostic; add algorithm families only via Extensions with explicit governing patterns and edition pins.
-
Anti‑pattern: “Shortlist = one winner.” Avoid: return the selected set even if it has one member. Choosing one among several selected candidates requires an explicit downstream decision rule.
-
Anti‑pattern: “Refresh plan described as prose triggers.” Avoid: record canonical
RSCRTriggerKindIdand payload pins; aliases only as labels and only if docked. -
Anti-pattern: “Packaging implies shipping governance.” Avoid: treat M5 as a library index; treat M6 as readiness wiring; ship only via
G.10.
G.1:9 - Consequences (informative)
- Repeatable authoring: CG‑Frame work becomes reconstructible: what exists, what it depends on, and how it is refreshed.
- Method pluralism with discipline: multiple generator/selector families can coexist without turning the chassis into a shadow method spec.
- Better reuse: outputs land directly in published artefacts (UTS/Name/RSCR‑ready) rather than remaining local notes.
- Lower refactor cost: method-wiring changes localise to Extensions; core invariants remain stable under their governing definitions.
G.1:10 - Rationale (informative)
- Six cards keep scope, harvesting, generation, selection, publication and refresh separately inspectable and changeable.
- Why “kit/index” rather than “pack”? A CG‑Frame authoring effort must stay modular; shipping is a separate governing boundary (
G.10). - Why put method-specific wiring into Extensions? It prevents conflating (i) universal invariants, (ii) frame‑specific kit surfaces, and (iii) method/generator families.
- Working models and micro-examples let authors check pins, names and boundaries before deeper formalisation.
G.1:11 - SoTA‑Echoing (informative)
The following practices connect to the chassis through its governed modules:
- Evidence synthesis: systematic and living reviews use explicit selection criteria and source-selection reporting; living review also needs an update policy. M2 obtains these through
G.2. PRISMA-style reporting can make the source selection inspectable. - Quality‑Diversity and archives: QD families (MAP‑Elites‑class, CMA‑ME‑class, and related archive‑based exploration) fit as M3/M4 extensions (
C.18/C.19) because they require explicit descriptor/distance/insertion pins and preserve set‑valued outcomes. - Open‑ended exploration: open-endedness systems (POET‑class, paired/adversarial environment generation lines, and curriculum-generation approaches) fit when treated as generator‑family wiring (governed elsewhere) rather than as chassis semantics.
- Set‑valued decision outputs: multi-objective and set‑valued evaluation practices align with the
G.Coreset‑return discipline, preventing hidden scalarisation. - Governed traceability: pinned policies and editions make the selected mechanisms and provenance inspectable across module boundaries.
G.1:12 - Relations
Builds on: G.Core, E.8, E.10, E.19.
Uses: A.10 (Provenance Anchors), A.15.2 (WorkPlan baselines), A.15.3 (planned fillings of independently declared positions), A.19.CN (CN‑Spec), G.0 (CG‑Spec), G.2 (SoTA Synthesis Pack), G.3 (CHR Pack@CG‑Frame), G.4 (CAL Pack@CG‑Frame), G.5 (Selector & Dispatch), G.10 (Shipping), G.11 (Refresh Orchestration), and (via Extensions) C.17, C.18, and C.19.
Publishes to / consumes from: Part‑F publication surfaces (UTS, naming, RSCR tests, Role/Concept artefacts) as cited by their governing definitions.
G.1:End
G.2 - Harvest and Synthesize SoTA for a CG-Frame
Type: Architectural (A) Status: Stable Normativity: Normative (unless marked informative)
Purpose. Provide a repeatable, auditable way to discover, triage, and synthesize state‑of‑the‑art (SoTA) across competing
Traditionlineages for CHR/CAL/LOG authoring and method selection in aCG‑Frame.Start here. Write the question the receiving CHR, CAL or selector work must answer. Before counting coverage, fix the source population and what counts as the same family. Distill the first source claims with their editions, evidence and limits, keeping competing lineages separate. The first useful result is a claim set that can answer part of that question and expose what is still missing. Use the manifest below when developing it into a conforming synthesis pack; a single-source fact lookup can return its source directly without creating such a pack. The primary output is a
SoTA Synthesis Pack@CG‑Framethat feeds:
- naming/publication (UTS),
- CHR authoring (G.3),
- CAL authoring (G.4),
- method/generator registries and dispatch (G.5).
Scope note. This pattern governs the harvesting + synthesis generator in Part G. Use G.10 to ship the pack and G.11 to orchestrate refresh.
Terminology (normative). The Tech token
Traditiondenotes a plural lineage with internally coherent commitments. Normative clauses may use the Plain term “tradition” only for that same meaning (1:1).
G.2:1 - Problem frame
A team extends FPF into a new CG‑Frame. The relevant literature is typically:
- plural (multiple
Traditionlineages with incompatible commitments), - source- and use-sensitive (results depend on the exact source and edition, claim region, EntityOfConcern, comparison basis, evidence, and receiving use),
- method‑heterogeneous (different evidence styles, operator sets, and validity regions),
- time-sensitive (new evidence or changed benchmarks and protocols can alter a claim’s applicability).
Downstream Part-G work in CHR, CAL, selection, shipping, and refresh depends on citation-ready claims that keep each exact CG-frame, source edition, claim region, EntityOfConcern, comparison basis, evidence anchor, and actual cross-source relation recoverable.
G.2:2 - Problem
How can we systematically assemble a SoTA view that is:
- pluralist but comparable (preserve plurality and establish the correspondence and comparison basis required for the intended use),
- evidence‑addressable (claims cite auditable evidence surfaces and anchors),
- actionable (supplies source-qualified claims and candidate inventories for G.3, G.4 and G.5),
- refreshable (the editions, policies and any applicable windows are pinned so RSCR/refresh can re‑audit and re‑run without semantic drift)?
G.2:3 - Forces
- Pluralism vs. consolidation. Consolidation is valuable, but unqualified fusion destroys meaning.
- Breadth vs. load‑bearing depth. Too broad becomes shallow; too deep misses rival lineages.
- Recency vs. stability. Freshness matters, yet durable “backbone” claims must be identified and kept visible.
- Pedagogy vs. rigour. Outputs must be teachable enough to support review, while remaining audit‑ready.
- Authoring vs. operations. This pattern governs authoring; use the applicable Work and decision patterns for operational runs and decisions.
G.2:4 - Solution
G.2:4.1 - G.Core linkage (normative)
Builds on: G.Core (Part‑G core invariants; citation/delegation hub)
GCoreLinkageManifest (normative).
(Canonical form, Nil‑elision, and Expansion rule are defined in G.Core.)
GCoreLinkageManifest := ⟨
CoreConformanceProfileIds := {
GCoreConformanceProfileId.PartG.AuthoringBase,
GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted
},
RSCRTriggerSetIds := {GCoreTriggerSetId.SoTAHarvestSynthesis},
CorePinSetIds := {GCorePinSetId.PartG.CrossingVisibilityPins}, // expands only for actual channel/receiving-use conditions under G.Core:4.2.3
CorePinsRequired := {
// Scope pins (G.2‑specific)
CGFrameId, // identifies the exact CG-frame, which is the declared framing episteme; its cited ClaimGraph keeps source and edition, claim regions, EntityOfConcern, comparison basis, and intended use recoverable
Tradition[],
entityOfConcern, GroundingHolon?, ReferencePlane, // claim subject, empirical ground when used, and plane
SoTA_SetId,
SoTAPaletteDescriptionId,
// Evidence / provenance pins (G.2‑specific)
CorpusLedgerId,
FlowRecordId,
EvidenceAnchorRef[],
EvidenceGraphId?,
// Crossing / synthesis pins (delta beyond CorePinSetIds; only when used)
GammaEpistSynthId[]?,
// Edition / policy pins (only when used)
HarvestPolicyRef?, // required when a coverage judgement is made
CoverageJudgementRef?, // the pack judgement, required for a relied-on coverage result
DistanceDefRef.edition?,
InclusionCriteriaId?,
ScreeningRubricId?
},
DefaultsConsumed := ∅,
TriggerAliasMapRef := ∅
⟩
(RSCR payload pins: ClaimSheetId[], SoTA_SetId, SoTAPaletteDescriptionId, BridgeMatrixId?, GammaEpistSynthId[]?, UTSRowId[]?, DistanceDefRef.edition?, HarvestPolicyRef?, InclusionCriteriaId?, ScreeningRubricId?, PathId/PathSliceId? when path‑citable evidence or a stable freshness window is pinned.)
Pattern‑local default. G.2 defines this coverage threshold outside the Part‑G-wide DefaultId set.
FamilyCoverageFloorK := 3 (unless explicitly overridden by HarvestPolicyRef and recorded in FlowRecord). This threshold supplies no counted population or same-family rule; those must be explicit before a coverage judgement. An undefined basis is unassessable, not a measured failure. Whenever coverage is judged, HarvestPolicyRef is required even when k uses this fallback; its applicability, receiving question, counted population/scope, grouping and same-family equivalence must be fixed before counting. An override changes k, not the unit or the independent pluralism duties.
Counted-family basis. The HarvestPolicy defines which candidates enter the counted population and when two entries represent the same family for this receiving question. Count equivalence classes under that rule. Repeated cards, aliases and source references for one family add zero. A combined method/generator population needs one receiving purpose and an overlap rule: a generator that is also a method is not counted twice unless the policy deliberately defines separate role-qualified units and justifies that interpretation. Freeze this basis before inspecting the count; changing it to turn a failure into three is not a repair of coverage.
The pack’s coverage judgement carries the policy/edition, counted units, deduplication basis, count, k and pass/fail result, or the exact missing basis when unassessable. Give this existing pack component a local CoverageJudgementRef for citation. Evaluate lineage and materially distinct entry plurality separately. Cards and downstream consumers cite this same judgement instead of choosing their own unit. Compare counts across packs only when their bases match, or after an explicitly justified common-basis recount.
G.2:4.2 - Kit: SoTA Synthesis Pack@CG‑Frame (surface governed by this pattern)
A conforming G.2 publication produces a notation‑independent pack whose internal organisation is free, but whose exported named components and views are stable and citable:
Each named component is addressable via a stable pack‑local identifier (e.g., CorpusLedgerId, ClaimSheetId, FlowRecordId) for citation and RSCR scoping. If any component is minted/evolved as a public id, it is published and cited via UTSRowId[] per CC‑GCORE‑UTS‑1 (delegation).
-
SoTA_Set@CG‑Frame(export view; “M2 output” consumed downstream) A read‑optimised view of the harvested candidate set for downstream generators and selectors. Constraint (normative):SoTA_Set@CG‑FrameMUST be reconstructible from pack components by id. Its coverage result cites the pack’sCoverageJudgementRef, including its fixed HarvestPolicy basis; the export view does not redefine family membership. -
G.2a CorpusLedgerLedger of candidate sources. Each row names the exact source and edition, claim region used, triage status (for example, include, park, or retire), evidence locator, and rationale for this CG-frame and receiving use. -
G.2b ClaimSheets[Tradition]Typed Claim Sheets perTradition, each with:- exact source and edition, claim region, effective ReferenceScheme where meaning matters, EntityOfConcern, and comparison basis for the stated use,
- explicit evidence anchors/citations (A.10 and/or EvidenceGraph refs when available),
- the currentness basis relevant to the claim’s use, including any applicable freshness window and risk/trust cues (cite
B.3when using trust/decay claims).
-
G.2c OperatorAndObjectInventoryInventory of candidate CHR terms (characteristics, scales and coordinates) and CAL operators or flows for downstream authoring. Cite the governing artefact and typing grounds for an entry already established as lawful. Mark the other candidates as stubs. -
G.2d BridgeMatrixA citable alignment/divergence surface acrossTradition×Tradition, with explicit losses and row scopes. If any row asserts substitution or fusion across sources or acrossTraditionrecords, the pack MUST attach aGammaEpistSynthIdrecord (alias:G.2‑F) perG.2:Ext.GammaEpistSynthesis(no silent fusion). -
G.2e MicroExamplesWorked micro-examples for load-bearing claims. Each names the exact source and edition, claim region, EntityOfConcern, comparison basis, and intended use; cites its evidence carrier or A.10 evidence-provenance path; and annotates the assurance contributions selected for that claim and use underB.3/B.3.3. When evidence-lane tags are used, cite their declared regime;C.2.2:4.6defines the KD-CALTA,VA, andLAlanes. The example card is only a publication form for those claims. -
G.2f UTSProposalsDraft Name Cards + Minimal Definitional Sheets (MDS) + alias proposals (incl. concept‑set linkage where applicable), with the required publication pins. -
G.2g entityOfConcern MapMap each key term, claim or public id to the entity or related targets its source claim concerns. Identify the empirical grounding holon when applicable, the reference plane and the cues needed to recover the claim for CHR/CAL authoring. -
G.2h PRISMA Flow RecordA screening/eligibility trail for how sources entered the pack (method‑profile is allowed; see Extensions). (Name is historical; the artefact remains notation‑independent.) The pack coverage judgement and its policy basis are recoverable here, separately from the lineage and material-entry pluralism results. -
G.2i SoSIndicatorFamiliesIndicator families as variants (windows/constraints/assumptions) with explicit Acceptance branches per variant (branch ids/labels only; threshold semantics belong to CAL governing definitions). -
G.2j MethodFamilyCardsCandidate method families with a shared signature and a plurality of implementations, each with validity regions, cost/complexity notes, and known failure modes. When the pack targets downstream registry/dispatch, MethodFamily cards SHOULD include the declared refs and pinsG.5needs (eligibility predicate refs, assurance profile cues, and the pack ids that justify the family). -
G.2k GeneratorFamilyCards(if applicable) Candidate generator families for environment/task generation with declared validity regions and transfer hooks. -
G.2l Annexes(optional; governing-definition-cited; see Extensions) For example: QD/NQD annexes, discipline‑specific indicator annexes, interop forms.
SoTAPaletteDescription (export view; required downstream)
Its pack-local SoTAPaletteDescriptionId identifies the view that binds these components.
- the
SoTA_Set@CG‑Frameview, ClaimSheetId[],OperatorAndObjectInventory,BridgeMatrixId?,SoSIndicatorFamilies(with variant/branch structure),MethodFamilyCards/GeneratorFamilyCards?,MicroExamples,UTSProposals,- and the
entityOfConcern Mapfor citation and later CHR/CAL authoring.
G.3/G.4/G.5 use this export view to find and cite the pack components.
Editorial template: 1‑page “SoTA Sheet” per Tradition (informative).
A one-page ClaimSheets[Tradition] template can collect the scope, claims, evidence anchors, validity region, failure modes, any applicable freshness window, cross‑Tradition reuse notes and pointers to micro‑examples.
G.2:4.3 - Harvester loop (conceptual choreography; pattern-governed)
Apply the following operations to the declared source population. A conforming pack satisfies its declared gates. When the available basis does not satisfy them, return the supported bounded result and the unmet condition; repeat a search or repair only when an attainable contribution justifies it.
-
Declare scope and plurality. Identify the exact CG-frame (the declared framing episteme), the initial
Traditionset, each intended claim region and EntityOfConcern, the comparison basis, and the receiving use. Record the cited CG-frame and source editions and evidence anchors in the pack pins rather than hiding them in a generic context field. Before counting, fix the HarvestPolicy’s receiving question, counted population, grouping and same-family equivalence, including overlap handling for a combined population. -
Discover and triage sources (ledger‑first). Populate
CorpusLedgervia:- adding seed sources,
- expansion via citation chaining and keyword family exploration,
- pruning using load‑bearing relevance tests tied to the declared CG‑Frame scope.
-
Distill claims per
Tradition. For eachTradition, author a Claim Sheet that preserves internal commitments and cites evidence anchors. Do not fuse cross‑Traditionclaims at this stage. -
Inventory operators/objects for downstream authoring. Identify candidate measurement terms and operators for downstream CHR/CAL authoring. Mark proposals as stubs. Cite an already governed CHR/CAL artefact and its grounds when reusing it. The inventory does not decide legality or set acceptance thresholds.
-
Build alignment/divergence surfaces. Where reuse across
Traditionis desired, record the obtaining correspondence and its exact basis inBridgeMatrix: F.9 for sense correspondence, C.3.3 for kind correspondence, or the direct rule for a plane relation, as actually used. State preserved distinctions and losses for the receiving question. Consolidation requires explicit alignment proof. Add bundle or gate anchors only for an independently applicable E.18 flow crossing or A.21 gate, underCC‑GCORE‑CROSS‑1. -
(Alias: G.2‑F) Construct a needed synthesis and record asserted fusion/substitution. Use B.1.3:4.4 to construct a joint conclusion from the source claims when the receiving question needs one. Recover their support and conflict relations, shared dependencies, subject and applicability before composing. A qualified synthesis with separately stated support can suffice when no common numerical model is justified.
If a
G.2pack publication asserts fusion or substitution across sources or acrossTraditionrecords (beyond mere “parallel divergent claims”), it MUST emitGammaEpistSynthIdrecords perG.2:Ext.GammaEpistSynthesis(provenance union + explicit object alignment refs + assurance tuple refs), and it MUST applyCC‑GCORE‑PEN‑1to its loss assessment and any numerical assurance penalty. -
Publish teachable micro‑groundings. Attach worked micro-examples that explain the load-bearing claim and a consequential variation in its conditions. Keep them within the claim’s supported scope. Each retains the source and edition, claim region, EntityOfConcern, comparison basis, intended use and evidence carrier or A.10 evidence-provenance path.
-
Apply gates and record repairs. Apply the fixed HarvestPolicy basis and count its distinct units before comparing coverage with
FamilyCoverageFloorK(and apply any optional diversity-by-distance gate under its own basis). Missing count semantics returns an unassessable result and the exact missing basis. If a defined gate fails:- record the failed criterion, count and basis in
FlowRecord, with the relevant corpus entries recoverable; - identify a missing relevant lineage, venue or source region that could change the result, and judge whether the available search or repair is worth undertaking for the receiving use;
- if it is, perform the search or repair and recount under the declared policy. Record what was done and the new result in
FlowRecord, citing any affectedCorpusLedgerentries. Pin the new references for any justified change to the policy or criteria and keep the earlier judgement with its original basis; - otherwise return the scoped claims and failed or incomplete pack judgement, its receiving-use limit and any precise remaining inquiry. Preserve the failed gate rather than treating the partial result as a conforming pack.
For example, an exhausted bounded population containing two method families fails a declared
k=3method-only gate. A further report about either family does not change that count. A known unexamined source region may justify another search; without one, the count and coverage failure can be the completed bounded answer. Missing population or equivalence rules instead leave the count judgement unassessable. - record the failed criterion, count and basis in
-
Emit hand‑off manifests and export views. Produce explicit manifests to:
so that downstream work can cite pack components by id rather than re‑authoring them. Each relied-on coverage result carries the same
CoverageJudgementRefand its policy basis; a downstream method-selection use cannot treat a combined method/generator count as a method-only count. The pack MUST also exportSoTA_Set@CG‑FrameandSoTAPaletteDescriptionas the default downstream consumption surfaces (ids pinned).
G.2:4.4 - Interfaces (minimal I/O Standard)
| Interface | Consumes | Produces |
|---|---|---|
| G.2-1 Harvest | exact CG-frame (the declared framing episteme) identified by CGFrameId, initial Tradition[], source edition and claim-region boundary, EntityOfConcern, comparison basis, receiving use, HarvestPolicyRef whenever coverage is judged | SoTA Synthesis Pack@CG-Frame (G.2a-G.2l) |
| G.2‑2 Extend | existing Pack + new sources/anchors + the applicable policy pins, updated if the policy changed | updated Pack + RSCR‑relevant trigger emissions (canonical kinds) |
| G.2‑3 HandOff | Pack | CHR‑handoff (to G.3), CAL‑handoff (to G.4), Registry‑handoff (to G.5) |
Note: Orchestration of re‑runs is governed by G.11; this pattern only defines what a conforming (re)harvest produces and what pins it must expose.
G.2:4.5 - Extensions (pattern‑scoped; non‑core)
Extensions are pattern‑scoped annexes. They do not introduce Part‑G‑wide norms; they declare the additional pins required when those semantics are active and cite the corresponding governing patterns.
G.2:4.5.1 - GPatternExtension: GammaEpistSynthesis
PatternScopeId: G.2:Ext.GammaEpistSynthesis
GPatternExtensionId: GammaEpistSynthesis
GPatternExtensionKind: GeneratorSpecific
GoverningPatternId: G.2
Uses: {G.Core, B.3, F.9, G.6} (penalty routing + trust/decay cues + bridges/CL + evidence path citation when used)
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
GammaEpistSynthId[](pack‑local ids of synthesis records; emitted iff fusion/substitution is asserted)EvidenceAnchorRef[](provenance union; evidence carriers cited by A.10 evidence-provenance paths)BridgeMatrixIdand the applicable correspondence references when a sense, kind or plane relation is used; includeBridgeCardId[]for an F.9 semantic bridgeCL/CL^planeandΦ/Ψ/Φ_plane policy-idswhen required by the cited crossing or actually used loss model (semantics and penalties →R_effremain governed by the cited definitions)PathId/PathSliceId?(only when citing viaG.6)
RSCRTriggerKindIds: {RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.ReferencePlaneEdit, RSCRTriggerKindId.PenaltyPolicyEdit, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.EditionPinChange}
Synthesis and its record (normative):
- For each asserted fusion/substitution, the auditable synthesis record identified by
GammaEpistSynthIdbinds: (i) provenance union, (ii) explicit object alignment refs, and (iii) assurance tuple refs underC.2.2:4.1, each tied to its exact assessed claim. A B.1.3Γ_epist^synthapplication and its returned episteme remain separate from this record. - This extension cites the
Γ‑fold,Φ, and penalty rules throughG.Coreand exposes the pins needed for replay. When B.3/C.2.2 supplies no justified common numerical score or loss calculation, retain the separate support, actual mapping limitations and bounded assurance conclusion; a synthesis record does not supply the missing model.
G.2:4.5.2 - GPatternExtension: HarvestProtocols
PatternScopeId: G.2:Ext.HarvestProtocols
GPatternExtensionId: HarvestProtocols
GPatternExtensionKind: MethodSpecific
GoverningPatternId: G.2
Uses: {B.3, A.10} (for freshness/decay and provenance anchors, when protocol requires them explicitly)
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
HarvestPolicyRef(declares the chosen protocol family and its parameters)FlowRecordId(protocol‑specific profile id or rubric id may be attached here)InclusionCriteriaId/ScreeningRubricId(ids only; semantics remain local to the protocol family)
RSCRTriggerKindIds: {RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.FreshnessOrDecayEvent}
Notes (extension discipline):
- This extension binds a declared protocol profile to the pack’s
FlowRecordwithout redefining evidence semantics.
G.2:4.5.3 - GPatternExtension: DHCAlignmentHooks
PatternScopeId: G.2:Ext.DHCAlignmentHooks
GPatternExtensionId: DHCAlignmentHooks
GPatternExtensionKind: DisciplineSpecific
GoverningPatternId: C.21 (DHC semantics)
Uses: {C.21, G.6, G.7} (DHC series + evidence path citations + bridge/CL regimes when alignment density is claimed)
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
DHCMethodRef.editionWindowRef?(if the DHC series is windowed)- exact F.17
SchemeSenseCellrefs used by the DHC comparison set (useSenseCellAddressRefwhere a durable address is needed; citeUTSRowId[]only for independently public ids) UTSRowId[]?(only if a cited cell or series id is independently minted or evolved as a public id)PathId[]/PathSliceId[](when alignment summaries cite evidence paths via G.6)
RSCRTriggerKindIds: {RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.TelemetryDelta}
Notes (extension discipline):
- If DHC alignment summaries are emitted, this extension ensures the DHC method edition and the cited evidence paths are visible.
- AlignmentDensity uses C.21’s Unit
obtaining_relations/100_compared_cells: fix the exact compared F.17 cell set and count the exact obtaining directed F.9 relations, retaining each relation’s orientation and admitted-use qualifier. Keep observed loss in its evidence account. A CL calibration label does not include or exclude a relation by itself. Any independently justified receiving-use filter must name its own policy and resulting population; it is not a C.21 CL threshold. - For example, three obtaining directed relations in a fixed set of 100 compared cells give a density of 3 in that Unit. Changing a CL label while relation truth, population and admitted-use qualifier remain fixed leaves the density 3. A fourth calibration row labelled CL=2 with no obtaining relation adds nothing. If a use condition actually changes which relations qualify, restate that changed population before comparing densities.
G.2:4.5.4 - GPatternExtension: NQDAnnex
PatternScopeId: G.2:Ext.NQDAnnex
GPatternExtensionId: NQDAnnex
GPatternExtensionKind: MethodSpecific
GoverningPatternId: C.18 (NQD-CAL semantics). C.19 governs explore/exploit logging when used.
Uses: {C.18, C.19}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
DescriptorMapRef.editionDistanceDefRef.editionInsertionPolicyRef(policy‑id/ref)EmitterPolicyRef(policy‑id/ref)TaskSignatureRef?(when QD mode is trait‑gated)
RSCRTriggerKindIds: {RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent}
Notes (extension discipline):
- This extension only pins the required references for replayability; it does not redefine QD semantics, dominance, or acceptance rules.
G.2:4.5.5 - GPatternExtension: InteropForms
PatternScopeId: G.2:Ext.InteropForms
GPatternExtensionId: InteropForms
GPatternExtensionKind: InteropSpecific
GoverningPatternId: G.13
Uses: {G.13}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
ExternalIndexRef.editionClaimMapperRef.editionMappingPolicyRef(policy‑id/ref)UTSRowId[](for published external ids/aliases where relevant)
RSCRTriggerKindIds: {RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TokenizationOrNameChange, RSCRTriggerKindId.EvidenceSurfaceEdit}
Notes (extension discipline):
- Interop affects only representation and citation routes; it must not introduce alternate legality gates or acceptance semantics.
G.2:4.6 - Palette first
SoTAPaletteDescription is the base, plurality-preserving palette. Keep it recoverable while displaying a derived front, archive or shortlist. A derived view declares its own comparison, retention or choice semantics; it does not replace the base palette or silently remove its members.
When the palette’s members are traditions, TraditionPalette is the reader-facing name for that same palette declaration. For methods, hypotheses or other members, use SoTAPaletteDescription or Palette + SubjectKind.
A TraditionFront is derived under a declared Q. A TraditionArchive is derived under a declared reachability or coverage rule. Pin each view’s governing basis separately; neither view renames Tradition or changes the base palette’s meaning.
Use richer geometry or atlas language when the comparison or retention question needs it to explain the derivation. Keep that interpretation separate from the palette’s default meaning.
G.2:4.7 - Optional atlas interpretation of a declared palette
Use TraditionAtlasView only when the reader needs several derived views or interpretive qualifiers together to understand a grouping, omission risk or comparison boundary. Otherwise use the palette and its declared front, archive or shortlist, or the thinner DeclaredSubstrateInterpretiveView. A naming-only question belongs to F.18.
TraditionAtlasView specializes DeclaredSubstrateAtlasView under A.19.DECLARED-SUBSTRATE-INTERPRETIVE-VIEW and retains that declaration by value: the base palette, active source set or result, TypedSetViews when several declared views are combined, the space/map references and interpretive qualifiers actually used, and the connections needed to answer the inspection question. Cite SearchSpaceRef or OutcomeSpaceRef for the corresponding declared spaces. Add SpaceMetricRef, TransitionRelationRef and BridgeDistortionNote only for the comparison, reachability, transition or cross-scale claim that needs them. An OutcomeMapRef identifies a mapping from the stated result into an outcome/effect space; it does not turn the palette or result into that space. Their formal claims retain their own governing definitions.
If the interpretation changes the base source-to-outcome relation or its distortion, reopen the substrate declaration. Different atlas views may use different spaces, metrics, relations or mathematical traditions; one view does not settle those choices for every other view.
G.2:5 - Archetypal Grounding (System / Episteme)
| Template element | U.System illustration | U.Episteme illustration |
|---|---|---|
| Tell | A safety engineering team needs to choose a control stack across robust-control, learning-based, and formal-verification lineages. It identifies the exact CG-frame (the declared framing episteme), vehicle and operating-envelope EntityOfConcern, source editions, claim regions, test or comparison basis, evidence anchors, and intended decision use. | A research group synthesizes SoTA on decision quality across named causal, evidential, bounded-rationality, and active-inference lineages, keeping each source edition, local claim, evidence norm, comparison basis, and intended research use explicit. |
| Show (failure) | The team merges source-local terms, treats incompatible test protocols and populations as comparable, and collapses partially ordered trade-offs into one unqualified score. A later safety review cannot recover which source, claim region, basis, or evidence supported the choice. | The group publishes one “best” metric and retrofits definitions to it. Conflicting claims cannot be traced because source editions, evidence anchors, comparison bases, and any actual cross-source relation were never made explicit. |
| Show (repair) | Keep parallel Claim Sheets with exact sources, editions, claim regions, EntitiesOfConcern, comparison bases, and evidence. Cite an F.9 Bridge and loss only for an actual relation. Authors of CHR, CAL, and selection methods can then use the citable claims without attributing authority to a card. | Preserve plural claims, represent indicators as families or variants, and expose freshness and evidence. Any justified alignment names its exact cells and obtaining relation; the card or matrix merely represents that result. |
G.2:5.1 - Count one pack under a declared basis
The following four hypothetical entries extend the control-stack example in §5. They illustrate how to count coverage under two different policies.
| Entry and distinct claim region | Lineage | Declared family unit |
|---|---|---|
| e1: robust controller’s operating-envelope claim | robust control | method M-R |
| e2: that family’s distinct disturbance-rejection claim | robust control | method M-R |
| e3: learned controller’s adaptation claim | learning-based control | method M-L |
| e4: scenario generator’s counterexample-generation claim | formal verification | generator G-S |
For the question “which control-method families can be selected?”, policy P-method counts method families only and equates entries exactly when they name the same declared method-family unit. The classes are {e1,e2} and {e3}: count 2, below k=3. Additional cards for e1 or citations for M-R leave the count at 2.
For the different question “which method and scenario-generator families can support building and evaluating this control stack?”, policy P-combined includes the three declared units M-R, M-L and G-S. Its overlap rule merges repeated references to one unit; these three units are stipulated distinct and G-S is not also M-R or M-L. Count 3 meets k=3 for that receiving purpose. This is not a passing method-only judgement and cannot be substituted after P-method fails. If one generator also qualified as a counted method, the overlap rule would have to resolve it before the count.
Four claim regions remain four material entries. The example independently has three lineages, so its two-lineage and three-material-entry pluralism duties pass under the stated facts in both policies. With no counted-family basis, family coverage is unassessable even though that pluralism result remains available. A justified k=2 override for P-method changes its threshold result while leaving its count at 2 and those independent duties unchanged.
G.1 M2 and the G.3–G.5 consumers cite the chosen pack judgement with its purpose and policy; they do not reconstruct a more convenient count from the cards.
G.2:6 - Bias-Annotation (informative)
-
Selection bias. Any harvesting protocol can over‑represent certain venues, languages, or evidence styles. Mitigation: pluralism floor + explicit
CorpusLedger+ explicit protocol pins. -
Consolidation bias. Pressure to “merge” lineages can erase incompatible commitments. Mitigation: keep Claim Sheets disjoint by default; require explicit alignment proof for fusion; preserve loss notes.
-
Recency bias. Overweighting newest papers can hide durable backbone results; underweighting them misses SoTA drift. Mitigation: state the search date and use-relevant currentness basis; make changes to relied-on sources or applicable windows RSCR-relevant.
-
Didactic bias. Micro‑examples can steer interpretation toward familiar domains. Mitigation: show the operation and a consequential variation within the claim’s supported scope, with explicit source anchors; add a different substrate only when the claim and its grounds cover that transfer.
G.2:7 - Conformance Checklist (normative) — CC‑G2
| ConformanceId | Requirement | Purpose / Notes |
|---|---|---|
| CC‑G2‑CoreRef | A conforming G.2 artefact MUST satisfy the effective core obligations declared by the GCoreLinkageManifest in G.2:4.1 (per G.Core Expansion rule). | Keeps core invariants governed by G.Core. |
| CC-G2-Pluralism-1 | A conforming pack MUST include at least two Tradition lineages and at least three materially distinct entries, each identified by its source edition and claim region, with its EntityOfConcern, evidence norm, and comparison limits visible. | Prevents a single lineage or one renamed source cut from masquerading as synthesis. |
| CC‑G2‑Ledger‑1 | A conforming pack MUST include G.2a CorpusLedger with inclusion/triage status and explicit rationale hooks per entry. | Makes discovery/triage auditable. |
| CC‑G2‑FlowRecord‑1 | A conforming pack MUST include G.2h FlowRecord that traces identification → screening → eligibility → included at a minimum granularity sufficient to reproduce the corpus boundary. | Prevents “mystery inclusion” and supports refresh. |
| CC-G2-ClaimSheets-1 | For each included Tradition, the pack MUST include a Claim Sheet identified by ClaimSheetId that names exact sources and editions, claim regions, effective schemes where meaning matters, EntitiesOfConcern, comparison bases, evidence anchors, freshness notes, and intended use; it MUST NOT fuse cross-Tradition claims by default. | Keeps plurality and provenance explicit without a Context container. |
| CC‑G2‑Palette‑1 | A conforming pack MUST export SoTA_Set@CG‑Frame and SoTAPaletteDescription as citable views (via SoTA_SetId, SoTAPaletteDescriptionId) and ensure both are reconstructible from pack components by id (no hidden extra structure). | Prevents downstream scraping of prose; keeps “M2 output” explicit. |
| CC‑G2‑Palette‑2 | If the pack exports one derived tradition view such as TraditionFront or TraditionArchive, it MUST keep SoTAPaletteDescription explicit as the default base palette, keep that derivation recoverable, and cite the declared Q or reachability/coverage rule that disciplined that view. Derived tradition views MUST NOT silently replace the palette’s default meaning. | Keeps non-default tradition views recoverable without redefining palette-first semantics. |
| CC‑G2‑AtlasInterpretation‑1 | If the pack exports TraditionAtlasView, it MUST satisfy §4.7 and the cited A.19 interpretive-view declaration by value, including the connections needed to answer the inspection question. | Keeps atlas use conditional and its interpretation recoverable. |
| CC‑G2‑entityOfConcernMap‑1 | A conforming pack MUST include G.2g entityOfConcern Map. For each load-bearing claim family and each minted or evolved public id, identify its subject or related targets, the applicable reference plane and any empirical grounding holon. Cite the relevant ClaimSheetId and evidence anchors (A.10 or G.6 paths when used). | Lets the reader recover what the claim concerns and what grounds it. |
| CC‑G2‑Alignment‑1 | Cross‑Tradition consolidation SHALL present either disjoint parallel claims with explicit divergence or an explicitly justified alignment proof. Reuse MUST cite the exact basis for every sense, kind or plane relation actually used and disclose its preservation and losses. Bundle/gate anchors MUST be supplied when an E.18 flow crossing or A.21 gate independently requires them, per CC‑GCORE‑CROSS‑1. | A kind correspondence alone requires neither an F.9 sense Bridge nor a flow-crossing bundle. |
| CC‑G2‑GammaSynth‑1 | If the pack asserts fusion or substitution across sources or across Tradition records (not merely “parallel divergent claims”), it MUST emit GammaEpistSynthId records satisfying G.2:Ext.GammaEpistSynthesis (provenance union + explicit alignment refs + assurance tuple refs). | Keeps the synthesis record (alias: G.2‑F) citable under its governing definitions. |
| CC‑G2‑Inventory‑1 | A conforming pack MUST include G.2c OperatorAndObjectInventory, sufficient for downstream CHR/CAL authoring to begin without re‑harvesting terms. | Ensures the pack is actionable. |
| CC‑G2‑Inventory‑2 | G.2c OperatorAndObjectInventory entries MUST NOT embed acceptance thresholds or decide legality locally. An entry may cite a governed CHR/CAL artefact with the grounds required by CC‑G2‑MeasurementLawful‑1. If an entry is not a citation of an already governed CHR/CAL artefact, it MUST be explicitly marked as stub (typing/lawfulness TBD) and MUST NOT be used as if lawful. Legality/threshold semantics are governed by G.3 for CHR and G.4 for CAL via explicit ids/pins. | Prevents “shadow CHR/CAL” and preserves lawfulness discipline without redefining it locally. |
| CC‑G2‑MeasurementLawful‑1 | If any inventory entry is presented as non‑stub (i.e., already lawful/typed), the pack MUST cite the governing lawfulness discipline (e.g., A.17–A.19/C.16 as applicable) and provide the minimal evidence anchors needed to justify that typing claim. | Prevents “quietly lawful” measurement claims inside the harvester pack. |
| CC-G2-MicroExamples-1 | For every load-bearing claim family, a conforming pack includes worked grounding that lets the reader obtain its result and handle a consequential change of conditions within the supported scope. Each example names the source and edition, claim region, EntityOfConcern, comparison basis and intended use; cites its evidence carrier or A.10 provenance path; and states the assurance contributions actually selected under B.3/B.3.3. | Preserve the grounding needed to understand and adapt the claimed contribution. |
| CC‑G2‑UTS‑1 | If the pack proposes or evolves any public ids, it MUST publish UTS proposals (Name Cards + MDS where applicable) and cite them via UTSRowId[], satisfying CC‑GCORE‑UTS‑1 (delegation). | Keeps naming and evolution disciplined. |
| CC‑G2‑Families‑1 | SoS indicators and candidate evaluation constructs SHALL be represented as families/variants (windows/constraints/assumptions) with explicit Acceptance branch structure per variant (branch ids/labels only), not as single unqualified scalars; any scalar summary MAY be included only as report‑only unless explicitly promoted by governing patterns. (Partial-order results retain their declared types under CC‑GCORE‑SET‑1.) | Prevents covert scalarization and keeps acceptance governed by downstream patterns. |
| CC‑G2‑HandOff‑1 | A conforming pack MUST emit hand‑off manifests to G.3, G.4, and G.5 that cite pack components by id and identify which families/operators are intended for downstream formalisation or registry entry. | Prevents downstream re‑authoring and drift. |
| CC‑G2‑CoverageGate‑1 | The pack MUST declare FamilyCoverageFloorK and enforce it as a harvesting gate. It MUST either (i) specify k in HarvestPolicyRef, or (ii) use the pattern‑local default rule governed by CC‑G2‑CoverageGate‑1. Default threshold (pattern-local): k=3. In both threshold branches, HarvestPolicyRef MUST fix the receiving question, population/scope, grouping, same-family equivalence and overlap rule before counting. The pack judgement cites that basis, its distinct units and result; all consumers reuse it. Missing basis returns unassessable. Duplicate references add zero, and a threshold override changes neither the units nor the independent duties in CC-G2-Pluralism-1. If the defined gate fails, record the criterion, count and basis in FlowRecord. When an identified relevant source region or operation can make a worthwhile contribution, repeat the search or repair and record that iteration and its result in FlowRecord. Otherwise return the bounded supported claims, failed/incomplete pack judgement and exact remaining inquiry under :4.3. A selected exploration policy is pinned. | Makes the coverage judgement and its receiving-use limit explicit. |
| CC‑G2‑DistanceGate‑1 | If a diversity‑by‑distance gate is used, the pack MUST pin DistanceDefRef.edition and the declared threshold (δ), and treat edits as RSCR‑relevant per CC‑GCORE‑TRIG‑* (delegation). If no such gate is used, the pack SHALL explicitly state that it is not used. | Avoids implicit distance defaults and improves refreshability. |
| CC‑G2‑RSCR‑1 | A conforming pack MUST emit canonical RSCRTriggerKindId causes (not free text) for edits to evidence surfaces, name/tokenization surfaces (e.g., UTS proposals/aliases), crossings, planes, edition pins, and harvesting policy pins (HarvestPolicyRef), per CC‑GCORE‑TRIG‑1…TRIG‑4 (delegation). | Keeps refresh reason codes stable and typed. |
| CC‑G2‑Ext‑GammaEpist‑1 | If G.2:Ext.GammaEpistSynthesis is used (i.e., any fusion/substitution is asserted), the pack SHALL expose the required pins listed in that extension and SHALL NOT redefine Γ‑fold/Φ/penalty semantics locally (cite governing definitions by delegation). | Keeps synthesis auditable without creating shadow specs. |
| CC‑G2‑Ext‑HarvestProtocols‑1 | If G.2:Ext.HarvestProtocols is used, the pack SHALL expose the required pins/criteria ids listed in that extension and SHALL NOT redefine evidence/quality semantics outside the declared protocol profile. | Keeps protocol variation explicit and separately citable. |
| CC-G2-Ext-DHC-1 | If G.2:Ext.DHCAlignmentHooks is used, expose the DHC method edition, exact compared F.17 cell population, counted obtaining directed F.9 relation refs with their orientation and admitted-use qualifiers, evidence paths, and C.21 Unit obtaining_relations/100_compared_cells. CL labels alone neither change that population nor impose a counting threshold. Cite any separate receiving-use policy under its own authority. | Keeps the quantity identical to the C.21/G.7 definition. |
| CC‑G2‑Ext‑NQD‑1 | If G.2:Ext.NQDAnnex is used, the pack SHALL expose the required pins/editions/policies listed in that extension and SHALL NOT redefine QD semantics locally. | Keeps QD/OEE extension pins replayable and non‑shadowing. |
| CC‑G2‑Ext‑Interop‑1 | If G.2:Ext.InteropForms is used, the pack SHALL expose the required interop pins and SHALL NOT introduce alternative legality/acceptance semantics. | Prevents “foreign gate” shadowing. |
G.2:8 - Common Anti‑Patterns and How to Avoid Them
-
AP‑G2‑1: “One true SoTA score.” Avoid: selecting a single unqualified scalar metric as “the” SoTA. Do instead: represent evaluation constructs as families or variants. Preserve the relationship established by each comparator and the candidates identified by any actual selection (G.Core).
-
AP‑G2‑2: Fusion without explicit alignment proof. Avoid: merging rival
Traditionclaims into one statement “by common sense.” Do instead: preserve parallel Claim Sheets; if consolidation is required, publish explicit alignment proof or keep a divergence record. -
AP‑G2‑3: Hidden protocol drift. Avoid: changing the harvesting protocol (inclusion criteria, windowing, screening rubric) without pins. Do instead: pin harvesting policy/profile ids and treat changes as RSCR‑relevant.
-
AP‑G2‑4: Unanchored pedagogy. Avoid: micro‑examples without carriers (they become folklore). Do instead: bind micro‑examples to A.10 anchors and declare
entityOfConcern. -
AP‑G2‑5: Atlas by default. Avoid: writing as if every tradition comparison or NQD/OEE note needs
TraditionAtlasView, or as if atlas wording renames the palette itself. Do instead: keep the base palette and derived front, archive, or shortlist explicit; use atlas form only when several declared views or interpretive qualifiers must be held together, and prefer thinnerDeclaredSubstrateInterpretiveViewwhen that is enough.
G.2:9 - Consequences
- Positive: Downstream CHR/CAL/dispatch work becomes faster and less ambiguous because the pack is citable and structured.
- Positive: Plural claims remain available, with the correspondences and comparison basis needed for any joint use.
- Positive: Refresh becomes tractable because pins and typed causes exist.
- Negative: Adds authoring overhead (ledger, flow record, micro‑examples, explicit pins).
- Negative: Requires governance discipline to prevent the pack from becoming an uncontrolled “everything bucket”.
G.2:10 - Rationale
Source-qualified claims let CHR/CAL authors and G.5 selectors reuse a comparison basis without re-harvesting it. Separate Claim Sheets preserve lineage commitments; justified mappings and support relations permit a qualified joint conclusion. The pack keeps these grounds citable for reuse and refresh.
G.2:11 - SoTA-Echoing — keep a reusable synthesis current without restarting it
Practice question. How should several downstream authors reuse a synthesis when new sources can change a consequential, unsettled comparison? The selected line keeps a question-bound source ledger, separable claims and explicit update causes; a living protocol is chosen when continuing evidence surveillance is worth its cost. A serious alternative is a well-reported static review with an explicit search date and a separately commissioned update. That alternative is sufficient for a stable question or a one-time decision and avoids maintaining a continuing review service.
The Cochrane Handbook v6.3 (2022), Chapter 22, §§22.2.3–4 supplies this earlier conditional living-review line and its resource trade-off: priority, uncertainty and likely new evidence can justify frequent updates, while additional searching needs resources. Adapt this conditional choice in HarvestProtocols; declare the search/update policy rather than treating every G.2 pack as perpetually living. The clinical-review guidance does not establish coverage or evidence adequacy for an arbitrary engineering CG-frame.
The PRISMA 2020 reporting guidance supplies a shared reporting basis for the static-review comparator. Its explanation of study selection distinguishes records, reports and included studies. Adapt the distinction to CorpusLedger, FlowRecord and the declared family units in §4.3: multiple source entries can concern one counted family. Reject treating a publication count or a completed flow diagram as proof of adequate family coverage. G.2’s k threshold, lineage duties and FPF alignment rules remain local decisions, not PRISMA requirements.
The 2024 PRISMA-LSR extension supplies the more specific reporting line for an actually selected living protocol: justify that mode, identify the version’s trigger and changes, and plan when to retire it. Adapt that distinction to the policy pins, change causes and downstream handoffs in §4.3 and G.2-2 Extend; G.11 governs refresh orchestration. Source reporting guidance does not supply an automatic update schedule or establish that a newly added claim is sound.
In the control-stack case (§5.1), if e1’s operating-envelope claim is revised, locate the changed source and claim in the ledger. Use the family card and handoff manifests to identify M-R and its affected consumers, then re-examine the claim and the comparisons that depend on it. The new report does not add a family. The method-only count remains 2 until a distinct method-family unit within the declared population is found. A static review can also be updated correctly, but each separate consumer must recover the changed basis unless a shared update is published. For the same new source and receiving question, the selected pack reuses existing screening and claim relations; its added maintenance burden is justified only while repeated shared use and consequential change make that work useful.
Reopen the choice of protocol when the evidence rate, decision importance, uncertainty or maintenance resources change; retire living surveillance when its reason no longer holds. Reopen an individual synthesis when a new source changes a relied-on claim, coverage judgement or alignment. A source with a new date but no relevant content change does not by itself warrant reconstructing every downstream conclusion.
G.2:12 - Relations
-
Builds on:
G.Core(core invariants, typed RSCR causes, Default Governing Definition Index)E.8(pattern template discipline)E.10(lexical/ontological rules; strict distinction; kind‑suffix discipline)E.19(conformance discipline)A.10(evidence-provenance paths and cited source/carrier anchors)A.19.DECLARED-SUBSTRATE-INTERPRETIVE-VIEW(generic interpretive-view and atlas discipline whenTraditionAtlasViewis used)A.6.P(space/view/publication precision restoration when palette/support claims collapse)B.3(trust, freshness/decay as cited governing patterns)F.9(bridges and CL as cited governing patterns)F.17(UTS publication discipline; via delegation)G.0(CG‑Spec legality gate; cited when legality surfaces are referenced)G.6(EvidenceGraph / path citation surfaces when used)
-
Used by:
G.1(generator chassis consumes harvested SoTA sets)G.3(CHR authoring consumes operator/object inventory and claim sheets)G.4(CAL authoring consumes operator stubs, acceptance branch scaffolding)G.5(registry/dispatch consumes MethodFamily/GeneratorFamily cards)G.10(shipping cites the pack as payload)G.11(refresh orchestration can re‑invoke harvest via typed causes)
-
Relates to:
G.2:End
G.3 - CHR Authoring for a CG‑Frame: Characteristics, Scales, Levels, Coordinates
Status: Stable
Tag. Architectural pattern (CHR kit; publishes lawful measurement primitives; constrains CAL authoring and selector/dispatch use)
Stage. design‑time (authoring & publication; enables admissible run-time consumption by G.4 / G.5)
Primary output. CHR Pack@CG‑Frame — a notation‑independent, UTS‑published CHR bundle that provides: typed Characteristics/Scales/Levels/Coordinates, legality + guard surfaces, aggregation/comparison specs, RSCR hooks/tests, and provenance pins.
G.3:1 - Problem frame
A team defining or revising a CG‑Frame (G.1) draws on competing SoTA traditions and constructs (G.2). Its CHR publication tells downstream users which measurements and operations the frame supports:
- CAL authoring (
G.4) needs typed, admissible operands and guard/legality surfaces to build admissibility and acceptance rules (thresholds and policy cut‑offs remain governed by CAL). - Selector/dispatch (
G.5) needs CHR‑typed quantities and explicit provenance pins so selection can remain set-returning and auditable under admissible orders. - Reuse beyond the defining source or use must name the exact characteristic and scale editions, bearer, scope and any temporal qualification required by the definition or receiving use, reference plane, evidence, and intended downstream use. Cite an
F.9relation only when it actually obtains between the namedF.17cells; any claim that relies on the relation for this use remains a separate bounded-use claim underC.2.1andF.9; its evidence-bearing reliance followsA.10and any applicableB.3assurance requirement.
The resulting publication is a CHR Pack that is CG‑Frame‑scoped, notation‑independent, and UTS‑published, with explicit edition/policy pins sufficient for reproducibility and RSCR.
G.3:2 - Problem
Numbers alone do not tell a downstream user which measurements and operations are admissible. Common failures include:
- Meaning leaks when the same token is reused after its referent or sense, bearer, scope, applicable validity window, evidence basis, or intended use has changed.
- Illicit arithmetic (e.g., averaging ordinals, mixing units, laundering polarity).
- Hidden normalizations that silently change scale type, polarity, or admissible transforms.
- Unreproducible comparisons (missing edition pins for methods/distances/policies; unclear reference plane).
- Unscoped reuse (the characteristic or scale lacks an exact edition, bearer, scope and any required temporal qualification, reference plane, evidence basis, or intended downstream use; any relation needed for reuse is left unstated).
- Un-auditable aggregation (no explicit legality surface and guard surface; no proof hooks; unclear governing definition for the Γ‑fold).
- Refresh chaos (changes in names/editions/policies do not map to typed RSCR causes).
G.3:3 - Forces
| Force | Tension |
|---|---|
| Pluralism vs comparability | Preserve tradition‑specific meaning ↔ enable admissible cross‑tradition use. |
| Expressiveness vs legality | Model rich measurement semantics ↔ block illegal operations “by construction”. |
| Portability vs honesty | Encourage reuse ↔ forbid implicit crossings and hidden loss. |
| Ease of authoring vs auditability | Keep authoring teachable ↔ require explicit pins, provenance, and tests. |
| Downstream flexibility vs upstream discipline | Let CAL/selector choose policies ↔ keep thresholds/policy cut‑offs out of CHR. |
G.3:4 - Solution — CHR authoring kit and publication surface
G.3:4.1 - G.Core linkage (normative)
Builds on: G.Core (Part‑G core invariants; citation/delegation hub)
GCoreLinkageManifest (normative).
GCoreLinkageManifest := ⟨
CoreConformanceProfileIds := {
GCoreConformanceProfileId.PartG.AuthoringBase,
GCoreConformanceProfileId.PartG.TriStateGuard,
GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted
},
CorePinSetIds := {
GCorePinSetId.PartG.AuthoringMinimal,
GCorePinSetId.PartG.CrossingVisibilityPins
},
// Pins strengthened for CHR authoring (delta over PinSets)
CorePinsRequired := {
// NOTE: the frame pin inherited from `GCorePinSetId.PartG.AuthoringMinimal` denotes only the exact
// declared CG-frame, which is the framing episteme here; it supplies no universal setting, scope, or reuse authority.
// Identify each measured bearer as the `entityOfConcern`. Inherit `CNSpecRef.edition` and `CGSpecRef.edition` when the kit or receiving use requires them. Each card adds its use-specific pins.
UTSRowId[], // required: CHR terms are public ids (Name Cards plus public-id continuity records)
PathId[]?/PathSliceId[]?, // when G.6 paths are used or required by the receiving contract
ReferencePlane, // required: definitional claims are plane-scoped
Φ/Ψ/Φ_plane policy-ids?, // when the applied loss/assurance model uses them or an applicable rule for the receiving use requires them (G.Core:4.2.3)
ΓFoldRef.edition? // iff an explicit Γ-fold artefact is pinned (otherwise use DefaultId)
// NOTE: method-/discipline-specific pins (e.g., DescriptorMapRef/DistanceDefRef/DHCMethodRef/InsertionPolicyRef)
// are declared only inside Extensions (e.g., `G.3:Ext.QD_OEE_Wiring`) to keep core linkage universal.
},
// consumed iff any published `CHR.AggregationSpec` relies on default Γ-fold (no explicit override pinned)
DefaultsConsumed := { DefaultId.GammaFoldForR_eff },
RSCRTriggerKindIds := {
RSCRTriggerKindId.EvidenceSurfaceEdit,
RSCRTriggerKindId.TokenizationOrNameChange,
RSCRTriggerKindId.CrossingBundleEdit,
RSCRTriggerKindId.ReferencePlaneEdit,
RSCRTriggerKindId.EditionPinChange,
RSCRTriggerKindId.PolicyPinChange,
RSCRTriggerKindId.DefaultGoverningDefinitionChange,
RSCRTriggerKindId.FreshnessOrDecayEvent,
RSCRTriggerKindId.LegalitySurfaceEdit,
RSCRTriggerKindId.BaselineBindingEdit
}
⟩
Apply G.Core:4.2 for conformance profiles, pins, nil-elision and expansion. The manifest names the applicable trigger kinds (G.Core:4.3) and uses G.Core:4.4 to locate the governing definition of each default it consumes.
G.3:4.2 - Output surface: CHR Pack@CG‑Frame (normative)
CHR Pack@CG‑Frame is the CHR kit payload that downstream patterns cite and pin. Its CN/CG references identify the governing specifications.
Minimum exported objects (kit surface):
CHR.Characteristic[]CHR.Scale[]CHR.Level[](when the scale type requires explicit level sets / order structure)CHR.Coordinate[](encodings + legality annotations; never an implicit “upgrade” of measurement structure)CHR.Guards(guard macro surface; semantics governed by cited definitions; seeG.CoreandA.18)CHR.LegalityMatrix(admissible operations per scale type / unit / polarity regimes)CHR.AggregationSpecs(typed aggregators/comparators + proof hooks + edition pins where applicable)UTSpublication bundle: Name Cards with designations under F.17/F.18, public-id continuity notes, and (when applicable) bridge and loss notes- RSCR artefacts:
RSCRTestId[], worked examples and their source/proof/measurement provenance, with applicable plane, path and policy pins
Mandatory provenance pins (conceptual, notation‑independent):
ReferencePlane- source/carrier references for worked examples and tests, including
PathId/PathSliceIdwhen G.6 is used or a receiving contract requires them - R-anchors to the supporting evidence under A.10. For evidence tagged under KD-CAL, retain its lane under
C.2.2:4.6. Use G.6 paths when shared path citation or the receiving contract requires them. - policy pins used by crossings or plane moves (when exercised)
- edition pins for any referenced method or metric definitions that affect interpretation
G.3:4.3 - CHR authoring chassis (S1–S8)
S1 — Charter the measurement scope (scope anchor).
If the needed characteristics are not established, use A.19.ECS:4.2 to derive them from the receiving use and consequential success/failure cases. G.0:4.5.1 connects that construction to the frame’s comparison and evidence requirements.
Identify the exact declared CG-frame (the framing episteme). For this CHR work, state the bearer or bearers and identify each as an entityOfConcern. Also state the scope, any applicable qualification and evaluation windows, ReferencePlane, evidence basis, intended downstream use, currentness conditions relevant to that use, and any contested expression whose reuse may require an actual F.9 relation. Output a design-time MeasurementCharter and KindMap.
When the receiving use depends on currentness, apply B.3.4 to the exact claim and use. Retain any binding calibration, regulatory or qualification window; use G.3:Ext.DecayWiring when pinning a currentness or decay model or policy.
If evidence-lane tags are used, declare the lane regime and its meanings here so downstream evidence discipline can remain lane-pure. B.3 and B.3.3 govern the assurance claim, receiving use and selected contributions; C.2.2:4.6 defines the KD-CAL TA/VA/LA lanes when that regime is selected. Evidence-path representation and audit remain governed by G.6; this pattern records wiring.
Lane docking (wiring‑only; normative).
If EvidenceLanes are used, the charter MUST:
- enumerate the lane tags used and cite the definitions of the selected regime, plus the upstream provenance for their use when available (e.g.,
SoTAPaletteDescriptionIdviaG.3:Ext.SoTAPackInputs); - expose any lane‑dependent tolerances / proof requirements via explicit pins (policy‑id and/or edition‑pinned refs), not prose;
- treat lane tags as provenance metadata (not Contexts): they MUST NOT be “bridged away” or silently mixed;
- for a cross-lane comparison or aggregation, cite the governing policy that permits this use (typically
G.4) and the grounds for the operation on these measurements. Retain G.6 paths when used or required.
If a crossing applies, use G.Core for its crossing rules and penalty routing.
S2 — Mint or reuse terms (UTS‑first). For each candidate characteristic, scale, level, or coordinate term: attempt reuse; otherwise mint via UTS Name Cards under F.17/F.18, with public-id continuity notes. When importing a term across contexts, identify its source-local meaning and the correspondence the receiving use relies on. Cite any required bridge and loss account.
S3 — Define CharacteristicCard (the per-characteristic publication unit).
A CharacteristicCard is the minimum unit CHR publishes for downstream legality. It SHOULD include (field names are indicative; semantics governed by cited definitions):
CharacteristicCard := ⟨
UTSRowId,
CharacteristicRef.edition,
entityOfConcern,
ClaimScope,
ApplicableSliceRef[],
QualificationWindow?,
EvaluationWindow?,
MeasurementMethodRef.edition?,
MeasurementModelRef.edition?,
EvidenceRef[],
IntendedDownstreamUse,
ReferencePlane,
ObjectKind,
Intent,
Definition (typed),
ObservableOf := ⟨instrument or determination protocol (cite its source or carrier under A.10), uncertainty treatment when it affects interpretation or use, temporal basis when consumed by the definition or receiving use⟩,
EvidenceLanes? (tags under the declared regime; supporting evidence refs; G.6 paths when used; cross-lane use under its cited acceptance/evidence policy),
ScaleRef.edition,
Polarity ∈ {↑, ↓, ⊥},
Domain/Range,
UnitSet,
Bounds / zero semantics (as applicable),
CurrentnessBasis? (for a use that depends on currentness; applicable window or decay model/policy under B.3.4),
Missingness semantics (typed; include a classification/mapping when non‑trivial; downstream tri‑state handling is per G.Core),
Stability/Reliability notes,
RoleDecls? := RoleDecl[] (wiring‑only; each role declaration names its governing pattern + required pins; see `G.3:4.5`),
QD.Role? ∈ {Q, D, QD-score} (interop alias for `RoleDecl` with `GoverningPatternId = C.18`; see `G.3:Ext.QD_OEE_Wiring`),
Micro-examples (source/proof/measurement anchors; G.6 paths and lane tags when applicable)
⟩
Actual C.16 measurement results retain the time stance, measurement Work and interpretation basis required by C.16:5.2–5.5.
Polarity gives the preferred direction for IntendedDownstreamUse: ↑ means higher-is-better, ↓ lower-is-better, and ⊥ no preferred direction assigned. Use ⊥ for a descriptive measurement. A target, range or other preference that has no single direction uses the applicable evaluation predicate and its Method; the Scale retains its measurement meaning under A.17/A.18.
Where RoleDecl := ⟨ roleLabel, GoverningPatternId, EditionPins?, PolicyPins? ⟩ (wiring-only; the value of GoverningPatternId names the FPF pattern that governs the role declaration semantics).
Rules (CHR‑governed intent, semantics governed by cited definitions where indicated):
- Scale/unit/polarity legality obligations cite MM‑CHR governing definitions (
A.18andC.16) and must be checkable by downstream patterns. - Missingness must be typed so downstream can apply tri‑state outcomes without silent coercion (tri‑state semantics are governed by
G.Core). - When recording
EvidenceLanes, use the lane tags defined by the selected regime. CiteG.6for evidence-path audits and the governing acceptance or evidence policy (typicallyG.4) for any permitted cross-lane use. - If
RoleDeclsare used, each declaration MUST cite the FPF pattern that governs the declaration, for exampleC.18orC.19, and surface the edition and policy pins required by that governing pattern; CHR does not define role semantics locally. - Role docking (normative, wiring-only): if any
RoleDeclis present withGoverningPatternId = X, thenG.3MUST include (or explicitly cite) a correspondingGPatternExtensionblock whose governing definition isX(or whoseUsesincludesX) and that surfaces the required pins for that role family. Otherwise the role declaration is non-conformant (it is an undocked semantic fragment). - Freshness docking (normative, wiring-only): if a characteristic’s freshness/half-life is defined via a named
decay model or policy, the relevant policy or reference MUST be pinned and cited through
B.3.4viaG.3:Ext.DecayWiring. - If a characteristic is intended to be promoted into
CG‑Spec, the linkage is explicit and edition‑pinned (wiring lives in an Extension; semantics governed byG.0).
S4 — Define ScaleCard and LevelCard (lawful measurement).
Publish the scale type and admissible transforms, plus levels or orders when applicable. Cite the MM‑CHR definitions that justify those operations for the frame’s characteristics.
Represent the distinctions needed by the measurement. A scale type may be combined with counting, exposure or cyclic constraints:
- Nominal / categorical: equality + counting; transforms are permutations.
- Ordinal: order‑preserving transforms; no arithmetic that presupposes intervals.
- Interval: positive affine transforms preserve order and ratios of differences. Differences are meaningful, and means may be lawful if justified.
- Ratio: positive scalar transforms; ratios meaningful; products/sums subject to unit discipline.
- Count / rates: explicit exposure/timebase requirements; rate conversions must be explicit.
- Cyclic: wrap‑around discipline + principal interval declaration.
S5 — Define CoordinatePolicy (encodings without hidden cardinalization).
When a numeric coordinate/embedding is used for convenience or tooling, CHR MUST publish:
- what invariants are preserved (order only / ratios / topology / wrap‑around),
- what remains illegal,
- what proof hooks are required if a structure with higher scale-type commitment is claimed.
Use the coordinate only for operations justified by the preserved invariants. A claim to stronger measurement structure needs the proof specified by the governing MM‑CHR definitions.
S6 — Publish legality + guard surfaces (Guard Macros + LegalityMatrix).
CHR publishes a CHR.LegalityMatrix and a CHR.Guards surface that downstream operators can reference.
Authors may give a guard macro a convenient name. Its definition MUST cite the governing rule for the check. The examples pair illustrative macro names with their governing sources:
CSLC_PROOF_REQUIRED(x)(MM‑CHR legality governing definitions:A.18/C.16)UNKNOWN_TRI_STATE(x)(tri‑state semantics governed byG.Core)UNIT_CHECK(x)(MM‑CHR legality governing definitions)RETURN_SET_FOR_PARTIAL_ORDERS()follows CC‑GCORE‑SET‑1. For A.19.CPM, retain the returned set of relation or poset tokens. A frontier under C.18:4.4 requires applying dominance to the stated candidate domain. A G.5 selection has its own outcome.METRIC_EDITION_REF(...)(edition‑pin discipline governed byG.Core; metric semantics governed byC.18/C.21as applicable)
S7 — Publish AggregationSpecs (typed, admissible, reproducible).
CHR may publish typed aggregation/comparison specs whose admissibility follows from their declared scales, constraints and proof basis, for use by G.4 and G.5. When deriving a characteristic of the whole from measurements of its parts, show that they concern compatible conditions and identify the joint dependency or bound supporting the aggregate (G.0:4.5.1). A direct measurement of the whole can supply the characteristic on its own measurement basis. For any published spec:
- The legality regime is explicit (scale/unit/polarity constraints + required proof hooks).
- If a contributor folding policy (Γ‑fold) is used and not explicitly overridden, cite
DefaultId.GammaFoldForR_effthroughG.Core.DefaultGoverningDefinitionIndex; do not restate the default here. - If method‑role declarations imply metric‑driven comparisons (e.g., QD roles), the relevant edition/policy pins are surfaced (wiring lives in an Extension; semantics governed by the referenced patterns).
S8 — Publish, test, and evolve (UTS + RSCR readiness). Publish the CHR pack and associated Name Cards to UTS. Attach:
- RSCR tests that check legality and guard coverage and reject illegal ops,
- worked examples with recoverable source, proof or measurement grounds; G.6 paths when used or required,
- applicable currentness and refresh conditions, plus deprecations with lexical continuity.
The pack supplies tests, examples and refresh conditions to the RSCR loop. G.11 governs its orchestration.
G.3:4.4 - Interfaces (normative)
| Interface | Consumes | Produces |
|---|---|---|
| G.3-1 Charter_CHR | exact declared CG-frame (the framing episteme), bearer or bearers identified as EntitiesOfConcern, scope and applicable windows, ReferencePlane, evidence basis, intended downstream use, and SoTA inputs (G.2) | MeasurementCharter, KindMap |
| G.3-2 MintOrReuse_Terms | candidate characteristic, scale, level, or coordinate expressions; their effective scheme, source-local sense, and intended CHR use; UTS registry | reused or minted ids and Name Cards where public ids are needed; exact F.17 cell refs and F.9 relation refs only when an actual relation between local meanings is required for the stated reuse and obtains |
| G.3‑3 Define_Characteristic | MeasurementCharter, candidate semantics | CHR.Characteristic[] (CharacteristicCards) |
| G.3‑4 Define_ScaleLevel | CharacteristicCard + MM‑CHR rules | CHR.Scale[], CHR.Level[] |
| G.3‑5 Define_CoordinatePolicy | Scale/Level + use‑case constraints | CHR.Coordinate[] + legality annotations |
| G.3‑6 Publish_GuardsAndLegality | Scale/Level/Coordinate set | CHR.Guards, CHR.LegalityMatrix |
| G.3‑7 Publish_AggregationSpecs | CHR set + legality hooks + (optional) metric refs | CHR.AggregationSpecs (+ proofs/refs + pins) |
| G.3‑8 Publish_CHRPack | all CHR artefacts + tests/examples | CHR Pack@CG‑Frame + UTS rows + RSCR tests |
G.3:4.5 - Extensions (pattern‑scoped; non‑core)
All blocks below are GPatternExtension modules (PatternScopeId-scoped; not new PatternIds). They store wiring only and cite governing patterns.
GPatternExtension: SuiteBoundaryLinkage
-
PatternScopeId:
G.3:Ext.SuiteBoundaryLinkage -
GPatternExtensionId:
SuiteBoundaryLinkage -
GPatternExtensionKind:
InteropSpecific -
GoverningPatternId:
A.19.CHR -
Uses:
{A.19.CHR, A.15.2, A.15.3} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum):
CHRMechanismSuiteDescriptionRef.edition?(when the suite description is cited as a reproducibility baseline)WorkPlanRefand local baseline locator (when a planned baseline binds CHR artefacts into WorkPlanning underA.15.2)CHRMechanismSuiteSlotFillingsPlanItemin that WorkPlan (when the plan fills an independently declared operation argument or relation position underA.15.3; cite the governing declaration and member designator, and locate the filling row within the plan)
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.BaselineBindingEdit, RSCRTriggerKindId.EditionPinChange} -
Notes (wiring-only):
A.19.CHRgoverns suite semantics and membership;A.19.CHR:4.1.2distinguishes ordinary planned baselines from typed filling.
GPatternExtension: SoTAPackInputs
-
PatternScopeId:
G.3:Ext.SoTAPackInputs -
GPatternExtensionId:
SoTAPackInputs -
GPatternExtensionKind:
DisciplineSpecific -
GoverningPatternId:
G.2 -
Uses:
{G.2} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum):
ClaimSheetId[]/ operator & object inventory refs (as cited inputs)SoTAPaletteDescriptionId?(when palette/traces are cited; used to dock contested‑term inventory and (if present) lane tags/tolerances)BridgeMatrixId?(when terms/constructs are imported across traditions)UTSRowId[]drafts/aliases from synthesis
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.TokenizationOrNameChange, RSCRTriggerKindId.CrossingBundleEdit} -
Notes (wiring‑only): SoTA pluralism inputs are governed by
G.2; this module only specifies which synthesis artefacts are cited while authoring CHR. When CHR authoring relies on pack coverage, cite the sameCoverageJudgementRefand HarvestPolicy basis, retaining the counted population and receiving question. CHR terms or cards do not redefine the counted family unit; lineage plurality remains a separate result.
GPatternExtension: CGSpecPromotionWiring
-
PatternScopeId:
G.3:Ext.CGSpecPromotionWiring -
GPatternExtensionId:
CGSpecPromotionWiring -
GPatternExtensionKind:
InteropSpecific -
GoverningPatternId:
G.0 -
Uses:
{G.0} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum):
CGSpecRef.edition(when a characteristic is promoted/linked intoCG‑Spec)CHR.Characteristic.idpointers included inCG‑Spec.Characteristics := [...](no shadow ids; CG‑Spec stores pointers, seeG.0)
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.LegalitySurfaceEdit, RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange} -
Notes (wiring‑only):
G.0governs promotion semantics and the legality gate; CHR only pins and cites.
GPatternExtension: MMCHRLegalityWiring
-
PatternScopeId:
G.3:Ext.MMCHRLegalityWiring -
GPatternExtensionId:
MMCHRLegalityWiring -
GPatternExtensionKind:
DisciplineSpecific -
GoverningPatternId:
A.18 -
Uses:
{A.17, A.18, C.16} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum):
- CSLC legality proof anchors/carriers (ids/refs as defined by MM‑CHR governing definitions; cite
A.18/C.16) - Unit coherence references (where units exist)
- CSLC legality proof anchors/carriers (ids/refs as defined by MM‑CHR governing definitions; cite
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.LegalitySurfaceEdit, RSCRTriggerKindId.ReferencePlaneEdit} -
Notes (wiring‑only): This module wires CHR artefacts to MM‑CHR legality proof obligations; legality semantics are governed by the referenced patterns.
GPatternExtension: DecayWiring
-
PatternScopeId:
G.3:Ext.DecayWiring -
GPatternExtensionId:
DecayWiring -
GPatternExtensionKind:
DisciplineSpecific -
GoverningPatternId:
B.3.4(freshness/decay semantics) -
Uses:
{B.3.4, G.6} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum):
- the currentness claim and receiving use, with any applicable
FreshnessWindowDeclRefor equivalent window pin DecayPolicyIdRef?(policy-bound; if decay model is referenced by id)- references to affected sources, carriers and any worked examples used to show drift. Include
PathSliceId[]when G.6 paths are used or required.
- the currentness claim and receiving use, with any applicable
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.FreshnessOrDecayEvent, RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.BaselineBindingEdit} -
Notes (wiring-only): Cite the B.3.4 currentness basis and any window/model actually used. A change in relied-on conditions or support can trigger refresh without assigning an expiry time or half-life.
GPatternExtension: QD_OEE_Wiring
-
PatternScopeId:
G.3:Ext.QD_OEE_Wiring -
GPatternExtensionId:
QD_OEE_Wiring -
GPatternExtensionKind:
MethodSpecific -
GoverningPatternId:
C.18 -
Uses:
{C.18, C.19} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum):
DescriptorMapRef.edition(if any Characteristic declares descriptor roles)DistanceDefRef.edition(if any Characteristic declares distance roles)DHCMethodRef.edition?(when a C.21 discipline-health measurement definition is actually used; a Q / QD-score role alone does not select it)InsertionPolicyRef?(when archive insertion semantics are declared for reproducibility)
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent} -
Notes (wiring‑only): QD/OEE semantics are governed by
C.18 and C.19. CHR only surfaces method‑role declarations (viaRoleDeclsor the interop aliasQD.Role) and the edition/policy pins required for reproducible archive/front interpretation.
G.3:5 - Archetypal Grounding
AG‑1 — ML fairness auditing.
System: deployed classifiers evaluated across cohorts in a CG‑Frame that makes abstention or deferral explicit.
CHR authoring: publish DemographicParityGap and EqualizedOddsGap as Characteristics with:
- explicit ReferencePlane, with deployment population and sampling regime recorded separately as scope and protocol conditions,
ObservableOf(audit protocol + uncertainty model + window),- interval scale (bounded; zero semantics explicit),
- missingness semantics (cohort sparsity and label noise are typed),
- legality surfaces and guard surfaces that forbid illicit cohort mixing and require explicit proof hooks for aggregation across cohorts.
Downstream: CAL acceptance binds thresholds and failure behavior; selector remains set‑returning under partial orders and may treat “defer/abstain” as a first‑class outcome (tri‑state semantics pinned through G.Core).
AG‑2 — Clinical diagnostics.
System: diagnostic pipelines compared in a CG‑Frame as datasets and protocols evolve.
CHR authoring: publish Sensitivity and Specificity as ratio‑scale, dimensionless Characteristics on [0,1], with:
- explicit
ObservableOf(trial protocol, inclusion criteria, uncertainty model), - freshness/decay expectations for the stated use,
- legality surfaces that forbid averaging incompatible ordinal labels (e.g., severity grades) and require explicit unit/exposure constraints for any derived rate.
Downstream: CAL acceptance governs thresholds and guard-bands; source/proof/measurement references, including G.6 paths when used, identify the basis to reconsider after a relevant change.
AG‑3 — Quality‑Diversity / Illumination (MAP-Elites/CMA-ME lineage).
System: a quality-diversity search system supplies an archive or front. In the CG-Frame, that set is the source for a declared set-result outcome under G.5.
CHR authoring: declare which Characteristics play Q/D/QD‑score roles and pin the metric definitions (descriptor map, distance definition, method editions) so archives are reproducible across runs and refresh can be triggered on edition changes. CHR does not scalarize partial orders; set‑return semantics are pinned through G.Core.
G.3:6 - Bias‑Annotation
CHR authoring is where many biases become “baked in” as measurement choices. Typical risks:
- Proxy bias: a convenient observable substitutes for the intended construct. Mitigation: require
ObservableOf+ ReferencePlane + micro‑examples; force explicit “what is being measured” rather than relying on labels. - Population and protocol shift: a change in sampling or protocol can alter the interpretation or applicability of a reported value, or the characteristic’s meaning when that procedure is part of its definition. Retain the population, protocol edition and use conditions, including any applicable validity or freshness window. Reconsider dependent claims when those conditions change. Use the corresponding RSCR trigger for a freshness/decay event or a change to the supporting evidence.
- Ordinal misuse bias: ordinal ratings treated as interval/ratio by convenience. Mitigation: publish scale type + admissible transforms; legality matrix + guard macros; reject coordinate upgrades without proof hooks.
- Cross-tradition meaning bias: an imported expression erases its source-local meaning or makes a changed bearer, scope, window, reference plane, evidence basis, or intended use disappear. Mitigation: name those values, cite exact
F.17cells and anF.9relation only when it obtains, and keep any downstream bounded-use claim explicit underC.2.1andF.9, with evidence-bearing reliance governed byA.10and any applicableB.3assurance requirement. Retain any mapping limitation and any justified penalty under the applicableG.Corerule. - Metric gaming bias (QD and evaluation): changing descriptors/distances changes what “diverse” means. Mitigation: edition‑pin metric definitions and make role declarations explicit (wiring via
C.18 and C.19).
G.3:7 - Conformance Checklist (normative)
| ConformanceId | Statement |
|---|---|
| CC‑G3‑CoreRef | G.3 is conformant only if the applicable G.Core obligations declared in G.3:4.1 are satisfied (effective expansion of profiles/sets + deltas; explicit pins; typed RSCR triggers; defaults with one governing definition). |
| CC‑G3‑01 | CHR Pack@CG‑Frame is published as a notation‑independent kit payload with the minimum exported objects listed in G.3:4.2. |
| CC-G3-02 | Every CHR.Characteristic names its exact characteristic and scale editions, its bearer as the entityOfConcern, claim scope and applicable slices, any applicable qualification and evaluation windows, ReferencePlane, method or model edition when it affects interpretation, evidence refs, intended downstream use, and a filled ObservableOf field (instrument or determination protocol, uncertainty treatment when it affects interpretation or use, and any temporal basis consumed by the definition or receiving use). |
| CC‑G3‑03 | Every CHR.Characteristic declares its ScaleRef, Polarity, and UnitSet (or an explicit “unitless” declaration), plus bounds/zero semantics where applicable. |
| CC‑G3‑04 | Missingness is typed in the CHR artefacts such that downstream tri‑state handling is possible without silent coercion. (Tri‑state semantics are governed by G.Core; the typing obligation is CHR‑local.) |
| CC‑G3‑05 | CHR.Scale / CHR.Level artefacts encode the scale type and admissible transforms, and make illicit arithmetic checkable by downstream consumers. |
| CC‑G3‑06 | Any published CHR.Coordinate includes a CoordinatePolicy that states preserved invariants and explicit non‑entitlements; coordinates do not silently upgrade measurement structure. |
| CC‑G3‑07 | CHR.LegalityMatrix and CHR.Guards exist and are referenced by downstream operator authoring; semantics are governed by cited definitions (MM‑CHR and G.Core), not duplicated locally. |
| CC‑G3‑08 | CHR.AggregationSpecs are typed and legality‑constrained; where Γ‑fold is required and no explicit override is pinned, cite DefaultId.GammaFoldForR_eff through G.Core.DefaultGoverningDefinitionIndex. |
| CC‑G3‑09 | If any characteristic is intended for promotion into CG‑Spec, the linkage is explicit and edition‑pinned (no shadow ids). (Governing definition: G.0; wiring via G.3:Ext.CGSpecPromotionWiring.) |
| CC‑G3‑10 | UTS Name Cards exist for public ids minted or evolved by the CHR pack (designations under F.17/F.18 plus public-id continuity notes). (Delegation target: CC‑GCORE‑UTS‑1 via CC‑G3‑CoreRef.) |
| CC‑G3‑11 | Worked examples and RSCR tests retain their source, proof or measurement basis, with G.6 path citations when used or required. They cover illegal-operation refusal, unit and scale constraints, polarity invariants and coordinate non-entitlements. |
| CC‑G3‑12 | Thresholds/guard‑bands are not embedded in CHR artefacts; they remain governed by CAL acceptance clauses (G.4). |
| CC‑G3‑13 | When method‑role declarations are present (via RoleDecls and/or QD.Role alias), each declaration is docked to its governing pattern via a corresponding G.3:Ext.* module, and the edition and policy pins required by the governing pattern are surfaced to make downstream interpretation reproducible. (QD/OEE governing patterns: C.18 and C.19; wiring via G.3:Ext.QD_OEE_Wiring.) |
| CC‑G3‑14 | Evidence wired. Each CHR.Characteristic cites the grounds of its legality and guard claims through A.10 source/carrier references, or G.6 paths when shared path citation or a receiving contract requires them. The chosen representation preserves the actual claim and support relations. |
| CC‑G3‑15 | An Archetypal Grounding section exists with at least two domain‑distinct examples that demonstrate lawful CHR typing/legality and the CHR↔CAL separation (notably: no thresholds in CHR). |
| CC‑G3‑16 | If EvidenceLanes are used, declare their tags with the selected regime’s definitions and pin any lane-dependent tolerances/proof requirements. Cross-lane comparison or aggregation requires both a governing policy that permits the use (typically G.4) and grounds for the operation on these measurements. Cite both, with G.6 paths when used or required. |
| CC‑G3‑17 | When CHR outputs are bound into a planned baseline, identify the A.15.2 WorkPlan and its local baseline locator. If the plan fills an independently declared operation argument or relation position under A.15.3, also use CHRMechanismSuiteSlotFillingsPlanItem with its governing declaration reference, member designator and plan-local filling-row locator, as governed by A.19.CHR:4.1.2 (wiring via G.3:Ext.SuiteBoundaryLinkage). |
| CC‑G3‑18 | Use-qualified currentness. A claim whose receiving use depends on currentness retains its B.3.4 basis and any applicable validity window or decay policy. Named models/policies are pinned through G.3:Ext.DecayWiring. Changes to relied-on windows, policies or other support conditions participate in RSCR under G.3:4.1. |
G.3:8 - Common Anti‑Patterns and How to Avoid Them
- Hidden cardinalization. Don’t treat ordinal encodings as interval/ratio; do publish coordinate policies that explicitly preserve order‑only invariants and forbid arithmetic upgrades.
- Unit laundering. Don’t add or average quantities with incompatible units; do force explicit unit discipline and legality checks via MM‑CHR governing definitions.
- Polarity drift. Don’t rely on “higher is better” implicitly; do publish polarity explicitly and make downstream use auditable.
- Threshold leakage into CHR. Don’t embed policy cut‑offs in CHR; do keep thresholds in CAL acceptance artefacts.
- Unpinned semantics. Don’t cite “the metric” or “the distance” without edition pins; do require edition‑pinned references when semantics affect interpretation.
- Unscoped reuse. Don’t reuse a characteristic or scale beyond its defining source and use on token continuity alone. Name the exact editions, bearer, scope and applicable window, reference plane, evidence, and intended downstream use; cite
F.9only when an actual relation between the named local senses is needed and obtains.
G.3:9 - Consequences
- Legality becomes checkable. Downstream patterns can reject illegal operations and rely on explicit legality surfaces rather than implicit conventions.
- Comparability without semantic flattening. Plural traditions remain representable because CHR preserves local meaning while making lawful relations explicit.
- Reproducible downstream behavior. Edition/policy pins make “why did this change?” answerable and RSCR actionable.
- Authoring overhead. The pattern shifts effort to up‑front authoring: explicit cards, pins, and tests are non‑optional when CHR becomes a public kit surface.
G.3:10 - Rationale
CHR authoring states which measurements and operations downstream reasoning may use. By making scale/unit/polarity explicit, publishing legality and guard surfaces, and requiring provenance pins, CHR authoring prevents downstream mechanisms from silently inventing their own legality assumptions.
Using shared rules from G.Core keeps tri-state guards, penalty routing, set-return semantics and RSCR trigger kinds consistent across Part G. Its default index points to the pattern that defines each default. CHR defines the frame’s measurement kit.
G.3:11 - SoTA‑Echoing
The pattern connects these evaluation practices to measurement authoring:
- treating abstention/defer and set‑valued outcomes as first‑class design objects (as in selective prediction and set-valued reporting),
- keeping multiobjective and archive‑based reasoning set‑returning rather than silently scalarizing (as in QD/illumination and open-ended evaluation),
- making evaluation semantics reproducible through explicit edition/policy pinning so users can recover how a reported metric was evaluated,
- modularizing method‑family specifics (QD/OEE, explore‑exploit) via explicit wiring and governing-definition assignment rather than embedding method semantics into universal measurement admissibility.
G.3:12 - Relations
Builds on: G.Core, G.1, G.2, G.6 when EvidenceGraph/Path citation is used or required, A.19.CHR, A.15.2 (WorkPlan baselines), A.15.3 (planned fillings of independently declared positions), A.17–A.18/C.16 (MM-CHR), F.0.1 (source-local meaning), F.1 (source selection), F.9 (actual relations between local-sense cells), F.17 (scheme-sense cells), F.18 (naming settlement), C.2.1 (bounded-use claims), B.3 / B.3.4, A.10, E.10, E.5.1–E.5.3.
Uses (via Extensions): G.0 (promotion/linkage to CG‑Spec), optional C.18 and C.19 (QD/OEE wiring).
Publishes to: G.4 (typed measurements, their permissible operations, any published aggregation or comparison specs, guard macros and applicable freshness pins), G.5 (role declarations and pins for reproducibility), UTS (Name Cards and public-id continuity notes), RSCR tests and hooks.
Constrains: any CAL/LOG/selector usage that consumes CHR (must treat CHR artefacts as typed/legal surfaces, not as prose hints).
G.3:End
G.4 - CAL Authoring for a CG-Frame: Operators, Acceptance Clauses, Evidence Wiring
Status: Stable
Use this when. A team has typed characteristics and now needs to publish reusable operators, acceptance clauses, and legal compositions before any candidate is evaluated. The working object is one design-time CAL Pack@CG-Frame, not an evaluation run, verdict, selector outcome, assurance case, or decision.
First move. Write one plain acceptance statement for one task: “For a subject x of the admitted kind, when the conditions of this use fall within claim scope S and satisfy any further temporal condition of the clause, apply operator O to a C.16 measurement result for Characteristic K that argument declaration R admits. When applying the clause, bind current result episteme E; the application of clause A returns pass | fail | unknown under threshold or policy P and any currentness conditions needed for that use.” Then turn only the reusable operator and clause definitions into stable CAL declarations. E belongs to the later application, not to reusable clause A.
Smallest viable CAL pack. Publish one charter for the exact CG frame, one typed operator card, one acceptance clause with ClaimScope, any evaluation window required by its predicate or receiving use, and unknown or failure behavior, one legal flow, one evidence and currentness profile, one proof-or-gap row and one worked declaration example. Publish a minimal editioned TaskMap that cites the exact charter, the C.22 TaskSignatureRef, and the declaration refs used by selection. Stop there when this pack answers the task; method-family extensions, archive surfaces, crossing records, and additional policy pins enter only when the case needs them.
What changes in practice. Thresholds and failure behavior become inspectable, illegal arithmetic becomes an authoring defect, and runtime workers can cite stable declarations when applying an operator or clause.
Not this pattern. Use C.16 for the measurement result, A.19 for comparison or selection, A.13 to identify each performer, A.15.1 for independent admission of dated evaluation Work, and F.6 when the receiving claim needs to establish that this Work was performed under the same assignment, A.6.1 for actual bindings, C.2.1 for the verdict episteme, A.10/G.6 for provenance, G.11 for currentness, B.3 for assurance, and C.11 for a decision. If the immediate question is whether the clause was applied and what verdict that application returned, use §4.4a to distinguish the declaration from its execution.
G.4:1 - Problem frame
CAL authoring starts from:
- one exact
CG‑Framewith itsEntityOfConcern,ReferencePlane, task, and assumption envelope, - a plurality of method traditions and claims (SoTA inputs), and
- CHR‑typed measurement constructs (
Characteristic/Scale/Coordinate+ legality guard macros).
Before downstream users run a selection, comparison, aggregation or selected-set formation, authors need an explicit, auditable CAL Pack that:
- defines what operators exist and what they are allowed to do over CHR types,
- externalizes fit-for-purpose acceptance as typed predicates whose use is bounded by an exact
ClaimScope, any evaluation window required by the predicate or receiving use, and any separate qualification window that limits use, and - binds these choices to an evidence wiring surface (lanes, provenance anchors, policy pins, and refresh triggers) so that downstream selection, logging, parity, and shipping can cite stable ids rather than re‑inventing semantics.
Use this kit to author and publish CAL declarations at design time. Use G.Core for shared Part G invariants, CG‑Spec for the comparison legality gate and CN‑Spec for governance.
G.4:2 - Problem
A CAL Pack connects typed measurements to later selection and evaluation. Drift or ambiguity in that connection can produce:
- Illicit operations include treating an encoding as cardinal without grounds, using incompatible units as if they agreed, and applying arithmetic that assumes intervals to ordinal values.
- Acceptance is scattered (thresholds embedded in code or in CHR prose; predicates not typed; unknown handling inconsistent).
- Evidence wiring is underspecified (which provenance anchors matter, what policy ids are in force, what is plane‑scoped, what changes must trigger refresh).
- Cross-sense or cross-plane imports are silent (hidden reuse across distinct source-local meanings, ReferencePlanes, or editions without the obtaining relation, required crossing records, and loss accounting).
- Tooling artifacts become semantics (vendor flags or implementation details substitute for a conceptual specification).
G.4:3 - Forces
- Expressiveness vs legality. CAL must allow useful comparisons/aggregations while staying lawful under CHR typing and legality gates.
- Pluralism vs comparability. Multiple method traditions must coexist without forcing premature unification, yet remain cross‑citable and auditable.
- Decision support vs auditability. CAL must support selection and selected-set formation while preserving explicit, reviewable assumptions and proofs.
- Exploration vs assurance. CAL must support exploratory regimes (probing, novelty, open‑ended search) without letting un‑assured outputs silently become dominance claims.
- Locality vs portability. Each CAL clause stays bounded by its declared
ClaimScope, any applicable temporal bounds, source meanings, and ReferencePlane; reuse beyond that boundary requires the exact relation and crossing records that the changed value calls for.
G.4:4 - Solution — author the smallest lawful CAL pack
G.4:4.0 - Practitioner authoring path C1–C9
Complete these actions in order. Extend a step only when the task needs the additional input named there.
- C1 — Charter the scope. Name the exact
CG‑Frame,EntityOfConcern,ReferencePlane, task,CNSpecRef.edition, andCGSpecRef.edition. State the assumption envelope in ordinary language. - C2 — Declare one typed operator. Give it a stable id, CHR-typed signature, preconditions, result kind, and failure behavior. The card declares the operation under
A.6.1. Evidence for an application comes from its execution (§4.4a). - C3 — Declare one acceptance clause. Name the exact Characteristic and the A.6.1 argument declaration that admits the corresponding C.16 measurement-result episteme. Declare the predicate or threshold,
ClaimScope, any evaluation window required by the predicate or receiving use, any separate qualification window that limits use, unknown handling, and the stated stop, degrade, or abstain behavior. Keep the exact current result episteme for the later application. If the clause claims statistical risk or coverage control, also name the loss, target, calibration population and window, sampling or exchangeability assumptions, declared treatment of shift, and the exact policy that states the guarantee. - C4 — Compose only a legal flow. Cite the operators and gating clauses. Preserve each operator’s declared result. For an admitted CPM comparison this is its relation-token set. When the flow uses a selector, preserve its declared outcome. A declared DAG describes a possible composition. Evidence of execution establishes which work ran.
- C5 — Name the minimum evidence/currentness need. Cite the exact A.10 source/provenance anchors and G.11 currentness conditions needed to judge the clause. Do not turn an evidence profile, citation, or graph membership into a verdict or actual reliance.
- C6 — Add an extension only when the task needs one. Select its current subject pattern first, then pin only the descriptor, distance, insertion, exploration, branch, or path records that change the present CAL action. Otherwise omit the extension.
- C7 — Record proof or an explicit gap. For every operator, flow, or clause, cite each required justification for legality, monotonicity or boundedness. If any required justification is missing, publish the gap and the consequent degrade/abstain behavior.
- C8 — Exercise declaration behavior. Provide one worked authoring example and focused conformance tests for illegal operations,
pass | fail | unknown, currentness conditions when they are part of admissibility, and failure behavior. The example and test remain declarations/test records unless separately grounded dated work is named. - C9 — Publish and hand off. Mint stable ids and continuity notes, then emit the smallest immutable
TaskMapedition. It cites the exact charter edition, the already constituted C.22TaskSignatureRef, the task, and edition-bearing operator, flow, gating-clause, and evidence-profile refs; the cited evidence profiles carry the needed currentness pins. It neither constructs the TaskSignature nor copies clause thresholds. Use G.11 for change refs; G.4 defines no refresh rule or runtime occurrence or result.
The path is complete when a reader unfamiliar with the author’s work can reconstruct the acceptance statement from the published ids and identify what must happen at runtime. The manifests, schemas, interfaces and optional extensions below make that same pack machine-citable. The order of work is C1–C9.
G.4:4.1 - G.Core linkage (normative)
Builds on: G.Core (Part-G core invariants)
GCoreLinkageManifest (normative). Canonical shape, Nil‑elision, and the Expansion rule are defined in G.Core.
GCoreLinkageManifest := ⟨
CoreConformanceProfileIds := {
GCoreConformanceProfileId.PartG.AuthoringBase,
GCoreConformanceProfileId.PartG.TriStateGuard,
GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted,
GCoreConformanceProfileId.PartG.ShippingBoundary
},
CorePinSetIds := {
GCorePinSetId.PartG.AuthoringMinimal,
GCorePinSetId.PartG.CrossingVisibilityPins
},
CorePinsRequired := {
UTSRowId[], // CAL artefacts are public ids (Name Cards plus public-id continuity notes)
ΓFoldRef.edition? // pin an actual numerical composition model/policy when used; otherwise cite the governing rule
},
// consumed iff no explicit `ΓFoldRef.edition` override is pinned
DefaultsConsumed := { DefaultId.GammaFoldForR_eff },
RSCRTriggerSetIds := { GCoreTriggerSetId.SoTAHarvestSynthesis },
RSCRTriggerKindIds := { // deltas (Expansion rule applies)
RSCRTriggerKindId.PenaltyPolicyEdit,
RSCRTriggerKindId.DefaultGoverningDefinitionChange,
RSCRTriggerKindId.BaselineBindingEdit
}
⟩
By the G.Core Expansion rule, the effective conformance ids / trigger kinds / pin obligations for G.4 are the expansions of the referenced profiles/sets/pin‑sets plus the explicit deltas above.
Notes (normative intent, delegated semantics):
- Apply
G.Corefor tri-state outcomes, penalty routing, set-return discipline, crossing visibility, the P2W split and typed RSCR causes. Its Default Governing Definition Index identifies the rule for each default. - EvidenceGraph/Path pins, when used, are declared only via
G.4:Ext.EvidenceGraphWiringin G.4:4.5. - Method‑specific pins (e.g., QD descriptor/distance/insert policy pins; open‑ended transfer rules pins) MUST appear only in Extensions blocks (see G.4:4.5) and MUST NOT introduce competing defaults.
G.4:4.2 - CAL Pack@CG-Frame surface (kit governed by this pattern)
CAL Pack@CG-Frame is the CG‑Frame’s published CAL Pack. Minimally, it provides:
-
CAL.Charter— identification and assumption basis for this CAL pack:- cites the exact
CGFrameId,EntityOfConcernRef, andReferencePlane, - cites the governance and legality records (
CNSpecRef,CGSpecRef) by edition, - records the assumption envelope on which the acceptance predicates rely without minting another governance or legality record.
- cites the exact
-
TaskMap— the conditional G.4 handoff record toG.5when CAL gates are current; one exact edition names the task, cites the already constituted C.22TaskSignatureRefand exact charter edition, and cites the acceptance-clause, operator, flow, and evidence-profile refs that selection consumes. It does not constitute the TaskSignature or contain threshold values. -
CAL.Operator[]— UTS‑published typed operation declarations governed byA.6.1; a card declares possible arguments, result kinds, and conditions but does not assert that an operation ran:- declared signature over CHR types,
- declared preconditions and postconditions, including references to legality guard macros,
- provenance and evidence hooks identified by ids or pins independently of tool behavior.
-
CAL.Acceptance[]— typed predicate declarations whose use is bounded by the declaredClaimScope, any evaluation window required by the predicate or receiving use, and any separate qualification window that limits use; a clause declares how an actual application is judged but is not itself a verdict:- binds to CHR characteristic ids and to exact A.6.1 argument declarations for admissible C.16 measurement-result epistemes (and, when inducing numeric comparison or aggregation, to
CG‑Spec.characteristicids), - keeps the exact current result episteme in the later application binding rather than in the reusable clause,
- exposes unknown handling and failure behavior via policy pins.
- binds to CHR characteristic ids and to exact A.6.1 argument declarations for admissible C.16 measurement-result epistemes (and, when inducing numeric comparison or aggregation, to
Each resultInputDeclarationRef resolves an A.6.1 ArgumentDeclaration. Its meaning identifies the C.16 measurement-result episteme expected by the clause. The declaration states the ValueKind and binding designation rule and requires a result for the named Characteristic with the admissible shape. A one-off clause may also cite an already existing episteme in fixedResultEpistemeRefs[]?; mark that clause one-off instead of presenting it as reusable.
-
CAL.Flow[]— legality‑checked declarations of possible operator composition; a declared DAG is not performed work:- declares the operator’s result kind, including a set of relation tokens for an admitted CPM comparison or the selected candidates for a set-returning selection. A scalar result requires a lawful operator,
- records which acceptance clauses gate which flows.
-
CAL.EvidenceProfiles— evidence wiring surface:- lane tags under their declared regime / provenance anchors / policy pins needed for
SCRand audit surfaces, - currentness conditions needed by the receiving use, with freshness-window and decay/Γ_time policy refs pinned only when those mechanisms are used.
- explicit
ReferencePlaneand any used penalty-policy refs (Φ(CL),Ψ(CL^k),Φ_plane); the ProofLedger supplies the receiving quantity, scale, input meanings, assumptions, and derivation or calibration. Monotonicity and boundedness alone do not justify a numerical loss.
- lane tags under their declared regime / provenance anchors / policy pins needed for
-
Optional
CAL.NQD[]— QD/OEE‑related calculus surfaces when declared:
A CAL clause that consumes G.7 calibration states its exact receiving use and premises. A locally justified CL threshold is an additional policy condition; the ProofLedger and named authority must support that condition. It does not replace the bounded-use claim, target classification or reliance result. A policy waiver changes only the authorized policy condition.
-
CAL.ProofLedger— a proof/justification ledger:- links operator/flow/clause ids to their needed legality and soundness results; a numerical support fold includes the B.3/C.2.2 receiving model, compatible scales and inputs, dependence assumptions, and boundary behavior.
-
Publication artifacts:
- UTS Name Cards under F.17/F.18 for all public ids,
- ids for RSCR tests and worked examples,
- deprecation notices and edition bump notes as public-id continuity records.
Boundary discipline (normative):
- No shadow specs: CAL artefacts cite
CN‑Spec/CG‑Specand do not introduce competing “local specs” (delegated; seeCC‑GCORE‑CN‑CG‑1via CC‑G4‑CoreRef). - Shipping boundary:
G.10governs shipping; seeCC-GCORE-SKP-1via CC-G4-CoreRef. - Refresh boundary: CAL publishes pins/payload for refresh;
G.11governs refresh orchestration.
Minimal schema fragments (notation‑independent; fields for citation, not an implementation schema):
CAL.Charter :=
⟨ charterId, charterEdition, cgFrameId, entityOfConcernRef, referencePlaneRef,
CNSpecRef.edition, CGSpecRef.edition, assumptionEnvelope ⟩
CALCharterRef := <charterId, charterEdition>
TaskMap :=
⟨ taskMapId, taskMapEdition, charterRef := CALCharterRef,
taskRef, taskSignatureRef := TaskSignatureRef,
acceptanceClauseRefs[], operatorRefs[], flowRefs[], evidenceProfileRefs[] ⟩
TaskMapRef := <taskMapId, taskMapEdition>
CAL.Pack@CG-Frame :=
⟨ calPackId, charterRef, taskMapRef, operatorIds[], acceptanceClauseIds[], flowIds[],
evidenceProfileIds[], proofLedgerId, nqdIds[]?,
utsRowIds[], workedExampleIds[], rscrTestIds[], publicIdContinuityNoteIds[] ⟩
CAL.Operator :=
⟨ operatorId(UTS), signature(CHR-typed), preconditions[], postconditions[],
evidenceProfileRefs[]?, failureBehaviorRef?, crossingRefs[]? ⟩
CAL.Acceptance :=
⟨ clauseId(UTS), characteristicRefs[], resultInputDeclarationRefs[],
fixedResultEpistemeRefs[]?, // deliberately one-off clause only
cgSpecCharacteristicRefs[]?, predicateRef, claimScopeRef,
evaluationWindow?, qualificationWindow?, unknownHandlingRef,
failureBehaviorRef, evidenceProfileRefs[]?, crossingRefs[]? ⟩
CAL.Flow :=
⟨ flowId(UTS), dag(operatorIds, edges), gateClauses(acceptanceClauseIds),
resultKind, decisionAidPolicyRef? ⟩
CAL.EvidenceProfile :=
⟨ evidenceProfileId(UTS), lanes(declared tags), anchors(A.10)[],
freshnessPolicyPins[]?, penaltyPolicyPins[]?, ΓFoldRef.edition? ⟩
lanes carries the tags and meanings of the declared evidence-lane regime; when a CHR charter supplies that regime, retain its tags and meanings. The assurance claim, receiving use and selected contributions follow B.3/B.3.3. Keep any consumed C.2.2:4.1 tuple, with its exact claim and F, G and R_eff values, separately recoverable from the lane tags.
CALCharterRef and TaskMapRef each resolve one immutable edition. A changed charter, task, TaskSignature edition, clause, operator, flow, evidence profile, or edition-bearing cited list creates a new taskMapEdition; an old TaskMapRef continues to resolve its old values. The C.22 TaskSignature remains a separately constituted episteme. The map relates that exact signature to the CAL declarations used by selection but neither derives the signature nor duplicates their thresholds.
When CAL authoring relies on SoTA coverage, consume G.2’s pack CoverageJudgementRef with its HarvestPolicy basis and receiving question. Preserve the distinction between family coverage and lineage/material-entry plurality. A CAL clause does not turn a combined method/generator count into method-only coverage.
G.4:4.4 - Interfaces (minimal I/O surface)
| Interface | Consumes | Produces |
|---|---|---|
G.4-1 Charter | exact CGFrameId, EntityOfConcernRef, ReferencePlane, CNSpecRef.edition, CGSpecRef.edition, assumption envelope, SoTA inputs, CHR Pack@CG-Frame | one immutable CAL.Charter edition and its CALCharterRef |
G.4-2 Operators | CHR typing + SoTA operator inventory | CAL.Operator[] (UTS ids; typed signatures; refs to evidence profiles & guards) |
G.4-3 Acceptance | task intent, exact Characteristic and A.6.1 result-input argument declarations, ClaimScope, any evaluation window required by the predicate or receiving use, any separate qualification window that limits use, policy pins, and CHR characteristics; exact result-episteme refs only for a clause explicitly marked one-off | CAL.Acceptance[] (typed predicate or threshold; admissible result-input declarations; scope; applicable evaluation and qualification windows; currentness pins when used; unknown and failure behavior refs) |
G.4-4 Flows | Operator cards + admissible aggregators | CAL.Flow[] (legality‑checked compositions; declared result kind) |
G.4-5 NQD Surface | Task intent + policy pins + (optional) QD/OEE inputs | CAL.NQD[] (descriptor/distance/insertion refs + edition pins; optional) |
G.4-6 Publish | all above, exact task, C.22 TaskSignatureRef, proofs, and examples | versioned CAL Pack@CG-Frame, exact CALCharterRef, and the smallest immutable TaskMap edition plus TaskMapRef, citing the task, matching TaskSignature, and acceptance-clause, operator, flow, and evidence-profile refs; also UTS entries, RSCR tests, Worked-Examples, and public-id continuity notes |
G.4:4.4a - Declaration-to-runtime evaluation boundary (normative)
A CAL pack is a reusable design-time declaration. A stored operator card, clause, flow, TaskMap, proof-ledger row, test, or evidence-profile reference establishes neither an actual participant nor performed evaluation. When a CAL declaration is applied, recover the following runtime chain:
- Name one exact
EvaluationMethod(U.Method). ItsU.MethodDescriptionmay state generic participants, parameters, effects, and evaluation conditions, but it carries no actual-participant slots and no intrinsic claim that a test, proof, or acceptance event occurred. - Cite the exact
CAL.Operator,CAL.Flow, andCAL.Acceptancedeclarations asA.6.1operation semantics. Resolve the clause’sresultInputDeclarationRef, then bind the exact current C.16 measurement-result episteme in this application and test it against the declaration’s Characteristic and admissible result shape. Use the exactA.6.1declaration and application bindings; do not infer them from a compatible signature,TaskMap, or stored reference. - First recover every precise performer’s A.13 core for the exact evaluation action, scope, working situation, and window, including the same obtaining assignment later used by any attribution. A.15.1 then independently admits one dated
EvaluationWork : U.Workfrom its performance history, enacted Method, extent, and containing-System relation. Add F.6 afterward only when the receiving claim needs exact assignment-bound attribution through that same assignment. Recover the evaluated or affected referent, actual resources, and every concrete participant through its direct subject relation or anA.6.1application binding. A compact attribution account may omit only an assignment identifier unused by its receiving claim; it omits no consumed fact. Ordinary activity not claimed asU.Workdoes not enter this branch. - State the local result under its direct predicate and pattern. A
CAL.Acceptanceapplication yields its exactpass | fail | unknownverdict; use A.19 for comparison and selection results, C.16 for measurement results, and C.11 for a decision result. No generic evaluation-result or work-result field substitutes for these objects. - When a durable assertion is needed, constitute one
C.2.1result episteme whose ClaimGraph states that local result, evaluated subject, interpretation basis, polarity or domain status, and uncertainty when current. The episteme is not the domain result and does not create it. - Attach source recovery and provenance through A.10/G.6 and currentness through G.11. State the A.10 evidence-provenance account and local
RelianceDispositionfor the bounded use, citing the direct rule for any reliance relation. Enter B.3 when an actual named assurance claim is current; a consequential use without one retains its direct governing rule. A citation, ledger edge, evidence profile, disposition, or assurance record does not establish the work, participant, application, or local result it describes. - A later selector, acceptance action, or decision is another governed occurrence. It relies on the result episteme through an exact premise, reference, decision-use, or operation-argument relation; mere storage, citation, or graph membership does not establish actual use.
This chain keeps declaration, execution, local result, result episteme, provenance, bounded reliance, currentness, acceptance, and decision independently recoverable.
G.4:4.5 - Extensions (pattern‑scoped; non‑core)
G.4 supports method‑family and discipline‑specific calculus variations exclusively via pattern‑scoped extensions.
GPatternExtension block: G.4:Ext.EvidenceGraphWiring
- PatternScopeId:
G.4:Ext.EvidenceGraphWiring - GPatternExtensionId:
EvidenceGraphWiring - GPatternExtensionKind:
InteropSpecific - GoverningPatternId:
G.6 - Entry: use when this CAL pack needs to cite an addressable G.6 path or slice, whether one receiving use requires it or several consumers share it.
- Stop: omit the block when a local A.10 source-to-use account is sufficient; remove it when no current clause, proof, or example cites the path.
- Uses:
{G.6} - ⊑/⊑⁺:
∅ - RequiredPins/EditionPins/PolicyPins (minimum):
EvidenceGraphId?PathId[]/PathSliceId[]UTSRowId[](for cited artifacts)
- RSCRTriggerSetIds:
∅ - RSCRTriggerKindIds:
{RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange} - Notes: CAL proofs and examples may cite G.6 evidence paths by these ids.
GPatternExtension block: G.4:Ext.NQD
- PatternScopeId:
G.4:Ext.NQD - GPatternExtensionId:
NQD - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
C.18 - Entry: use only when the current task applies a C.18 quality-diversity/archive method and its descriptor, distance, insertion, or archive policy must be pinned for CAL use.
- Stop: omit or retire the block when the task has no current archive/QD clause or when those refs no longer change a CAL action.
- Uses:
{C.18} - ⊑/⊑⁺:
∅ - RequiredPins/EditionPins/PolicyPins (minimum):
DescriptorMapRef.editionDistanceDefRef.editionInsertionPolicyRefArchiveRef?TaskSignatureRef?(if activation is TaskSignature‑bound)
- RSCRTriggerSetIds:
∅ - RSCRTriggerKindIds:
{RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent} - Notes: C.18 defines QD semantics. CAL pins the descriptor, distance and insertion records needed to reproduce archive behavior. Any archive/illumination summaries (e.g., coverage / QD‑score / occupancyEntropy / filledCells) are published as report‑only outputs unless an explicit CAL acceptance clause/policy authorizes promotion.
GPatternExtension block: G.4:Ext.EELog
- PatternScopeId:
G.4:Ext.EELog - GPatternExtensionId:
EELog - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
C.19 - Entry: use only when the current task has a C.19-governed exploration/exploitation budget or probe-accounting rule that changes a CAL clause or failure branch.
- Stop: omit or retire the block when no current CAL action consumes those C.19 refs.
- Uses:
{C.19} - ⊑/⊑⁺:
∅ - RequiredPins/EditionPins/PolicyPins (minimum):
ExploreExploitBudgetPolicyRefProbeAccountingRef?FailureBehaviorRef?(if probe/sandbox is policy‑bound)
- RSCRTriggerSetIds:
∅ - RSCRTriggerKindIds:
{RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent}
GPatternExtension block: G.4:Ext.SoSLogBranches
- PatternScopeId:
G.4:Ext.SoSLogBranches - GPatternExtensionId:
SoSLogBranches - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
C.23 - Entry: use only when C.23-governed SoS-LOG branches currently explain a CAL degrade/abstain path.
- Stop: omit or retire the block when those branch/rule ids no longer change a current CAL clause, flow, or explanation.
- Uses:
{C.23} - ⊑/⊑⁺:
∅ - RequiredPins/EditionPins/PolicyPins (minimum):
SoSLogRuleId[]SoSLogBranchId[]FailureBehaviorPolicyId
- RSCRTriggerSetIds:
∅ - RSCRTriggerKindIds:
{RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.MaturityRungChange, RSCRTriggerKindId.TelemetryDelta} - Notes: The pinned C.23 branches and rules explain the clause’s degrade/abstain behavior.
G.4:5 - Archetypal Grounding
Tell. A team working within a CG‑Frame must choose and justify a set of candidate methods (possibly a selected set or a set retained in an archive) under explicit legality, evidence, and scope constraints. CHR provides the typed measurement basis; CAL declares auditable predicates and flows that separately grounded runtime work may apply.
Show 1 (bounded CAL pack skeleton).
Use: R&D selected-set choice. The pack names the exact CG frame and EntityOfConcern, candidate-set ClaimScope, ReferencePlane, and evaluation window. CHR defines SafetyClass(ord↑), CostUSD_2026(ratio↓), Readiness(nominal).
CAL.Operator: DominatesParetoSignature over CHR types, precondition references CHR guard macros.CAL.AcceptanceClause: AC_SafetyGateReusable typed predicate forSafetyClass(and its levels), citingSafetyResultArgument-D1, the A.6.1 argument declaration that admits a C.16 measurement-result episteme for that Characteristic. Thresholds are valid for the statedClaimScopeand evaluation window; unknown handling uses tri-state pins. The clause names no current measurement-result episteme.CAL.Flow: Flow_ParetoPortfolioDeclares a selected-set result kind; gated byAC_SafetyGateandAC_Budget.CAL.EvidenceProfile: EP_SafetyEvidenceDeclares anchor ids and freshness policy pins required forSCR.
When this CAL pack supplies selector gates, it publishes TaskMapRef=<SafetySelectionMap, E3>. That map cites CALCharterRef=<SafetyCALCharter, E2>, C.22 TaskSignatureRef=SafetyPortfolioTaskSignature-E4, the exact task, and edition-bearing refs AC_SafetyGate-E2, Flow_ParetoPortfolio-E1, DominatesPareto-E3, and EP_SafetyEvidence-E4; it contains no threshold values. Downstream G.5 consumes the exact TaskSignatureRef and this TaskMapRef together, verifies that the map cites the same signature, and resolves the charter and declarations through their refs. If the charter or a cited clause changes, G.4 publishes another TaskMap edition; selectors that still cite <SafetySelectionMap, E3> continue to replay the old boundary.
Show 2 (explicit cross-sense or ReferencePlane import).
A SafetyClass result uses an expression with a different F.17 source-local meaning or comes from another ReferencePlane. CAL may author a clause using it only after the exact F.17 cells and obtaining F.9 relation are cited when meanings differ, and the applicable plane or edition crossing records are cited when those values differ. The clause keeps its declared ClaimScope and window; the import does not silently widen either.
Show 3 (one performed acceptance evaluation).
Before the candidate action is admitted as Work, A.13 recovers SafetyEvaluatorSystem-17 : U.System for exact action SafetyAcceptanceEvaluationAction-17. Its admitted SafetyEvaluatorBoundary-17 contains the evaluation controller, its active decision state, and the input/output channels through which it applies the clause; it excludes the CAL declarations, measurement-result episteme, candidate, assignment, and containing team System. The action’s scope is SafetyAcceptanceClaimScope-17, its working situation is SafetyGateEvaluationSituation-17, and its window is 2026-07-30T09:00:00Z through 2026-07-30T09:20:00Z. It is directed by SafetyAcceptanceDecisionNorm-17: apply the current clause to admissible current inputs, return unknown rather than force a threshold verdict when uncertainty crosses the boundary, and reject an input whose declared result shape is incompatible. The relevant conditions are clause edition, result-shape admissibility, measurement currentness, and uncertainty relative to the threshold.
The local kind SafetyAcceptanceEvaluatorSystemRole is declared under A.2. Its membership criterion requires the stable work-facing contribution of safety-acceptance evaluation and goal-directed, condition-sensitive regulation under SafetyAcceptanceDecisionNorm-17: the holder must bind admissible inputs, choose the clause-defined verdict, and abstain or return unknown when the declared conditions require it. SafetyEvaluatorDecisionTrace-17 shows SafetyEvaluatorSystem-17 rejecting an incompatible result shape and returning unknown when the admissible uncertainty interval crosses the threshold; the system-boundary and runtime records show that those actions occurred within SafetyEvaluatorBoundary-17. The cited decision-trace, system-boundary and runtime records support the criterion facts under A.2’s membership rule; A.10 makes that source-to-use account recoverable. The case independently classifies SafetyEvaluatorSystem-17 under SafetyAcceptanceEvaluatorSystemRole. Neither the candidate Work nor the assignment supplies that classification. No Grade, autonomy result, characteristic profile, or stronger assurance claim is consumed here.
The same A.13 core uses SafetyAcceptanceEvaluationAssignment, a directly declared U.SystemRoleAssignment species under A.2.1. The species defines holder, assigned-kind, and evaluation-candidate participant meanings; its predicate appoints the holder to evaluate that candidate under the applicable clause for the stated scope, situation, and window. SafetyAcceptanceEvaluationAssignment-17 obtains with SafetyEvaluatorSystem-17 as holder, SafetyAcceptanceEvaluatorSystemRole as assigned-kind value, and C-17 as evaluation candidate. Its maximal uninterrupted predicate-true interval covers the full stated window.
Only after that A.13 core is established does A.15.1 independently admit EvalWork-2026-07-30-17 : U.Work from its exact performance history, enacted SafetyAcceptanceMethod, temporal extent, and obtaining containing-System relation to independently admitted SafetyEvaluationTeamSystem-17. The actual A.6.1 application SafetyAcceptanceApplication-17 binds candidate C-17 separately from its binding of current C.16 measurement-result episteme SafetyMeasureResult-E17 to SafetyResultArgument-D1 while using unchanged reusable clause AC_SafetyGate. Neither the assignment nor an F.6 conclusion is an A.15.1 admission premise.
Because this worked case states that the Work was performed under an assignment, F.6 afterward establishes performedUnderAssignment(EvalWork-2026-07-30-17, SafetyAcceptanceEvaluationAssignment-17) through the same obtaining A.13 assignment. The direct case fact links the named Work and assignment, the holder is the performer, and the assignment interval covers the Work. A different overlapping assignment held by the same performer would not establish this attribution.
SafetyMeasureResult-E17 states the measured safety characteristic, scale, attributed value, uncertainty, model, calibration, and measurement Work; it is neither the raw detector output nor the acceptance verdict. The clause application obtains unknown because the uncertainty interval crosses the threshold. A later SafetyMeasureResult-E18 can bind through separately identified SafetyAcceptanceApplication-18 while AC_SafetyGate remains unchanged. A C.16 result for CostUSD_2026, a result with an incompatible declared shape, or raw detector output fails SafetyResultArgument-D1 before the predicate runs; it does not cause a new reusable clause edition. A separate C.2.1 episteme asserts that exact verdict and cites its provenance under A.10 and, when the EvidenceGraph extension is present, G.6; G.11 supplies currentness. A later C.11 result may record defer, and its claim uses the verdict episteme through an exact premise or decision-use relation. Any decision-making Work remains separate. The clause card, proof-ledger row, evidence edge, and decision record do not retroactively establish the measurement Work or the evaluation occurrence.
Show 4 (a support model and its separate acceptance threshold).
For the pump-triage use in G.5 §0.5, consider a separately declared gated variant. AC_InputConditionGate-E1 accepts a declared probability of at least 0.85 that both the measurement series is valid (event A) and its time alignment is valid (event B) for the same 24-hour input. Its scope, evaluation window, result-input declaration, and unknown behavior belong to this G.4 clause, not to the default composition rule. PumpInputEvidence-E1 supplies the receiving model and provenance for those event meanings and probabilities. An applicable TaskMapRef binds that clause, profile, and the matching C.22 task signature.
Suppose the application inputs establish P(A)=0.9 and P(B)=0.9 and the profile’s model establishes independence. The model gives P(A ∩ B)=0.81, so the clause returns fail. Minimum 0.9 would incorrectly pass. If independence or a justified conditional alternative is unavailable, the missing joint probability yields the clause’s unknown result, with the declared downstream degrade/abstain behavior; a high F or formal line tag cannot replace it.
For a different receiving question that asks only for a qualified comparison on existing heterogeneous support, the profile retains the proof, empirical contributions, shared-bias limitations, and contrary evidence under B.1.3/C.2.2. It does not invent a probability or apply AC_InputConditionGate-E1 to that different question. The ordinary G.5 shortlist remains available without an additional assurance calculation.
G.4:6 - Bias-Annotation
CAL clauses define what counts as acceptable. The choices below can bias that judgement:
- threshold‑selection bias (arbitrary floors masquerading as natural laws),
- measurement bias amplified by illegitimate arithmetic or hidden scalarization,
- survivorship bias in Worked‑Examples and probe telemetry,
- Goodhart pressures when report‑only telemetry is accidentally treated as dominance.
The pattern mitigates these by requiring typed acceptance clauses, explicit policy pins, and an auditable ledger of proofs and justifications, while identifying cross-sense and ReferencePlane reuse. For a numerical assurance penalty, the model for the receiving use determines which results it affects.
G.4:7 - Conformance Checklist (normative)
| ConformanceId | Statement |
|---|---|
| CC‑G4‑CoreRef | Conformance with G.4 requires satisfying the effective G.Core obligations referenced by the GCoreLinkageManifest in G.4:4.1 (profiles, pin sets, consumed defaults, and trigger kinds). |
| CC‑G4‑01 | CAL Pack@CG-Frame is published as a notation-independent object with stable UTS ids (Name Cards under F.17/F.18) for CAL.Charter, TaskMap, all operator, acceptance-clause, flow, and evidence-profile declarations, Worked-Examples, and public-id continuity notes, including deprecations and lexical-continuity notes. Tooling/vendor details remain non-normative. |
| CC‑G4‑02 | Each exact CALCharterRef = <charterId, charterEdition> resolves one immutable charter edition naming the exact CGFrameId, EntityOfConcernRef, ReferencePlane, CNSpecRef.edition, CGSpecRef.edition, and assumption envelope on which the pack relies. |
| CC‑G4‑03 | Every CAL.Operator has a declared CHR‑typed signature and stated preconditions. Cite each referenced legality guard macro by id. |
| CC‑G4‑04 | Every reusable CAL.Acceptance binds the exact Characteristic and exact A.6.1 resultInputDeclarationRef values that declare the admissible C.16 measurement-result episteme inputs; the exact current result episteme is bound only in an actual application. A clause marked one-off may additionally cite fixedResultEpistemeRefs[]?. Every clause also declares its predicate or threshold, ClaimScope, any evaluation window required by the predicate or receiving use, any separate qualification window that limits use, unknown handling, and failure behavior. A statistically risk-controlled clause also names its loss, target, calibration population and window, sampling or exchangeability assumptions, declared treatment of shift, and the exact policy that states or defines the guarantee. Inputs with distinct source-local meanings cite the exact F.17 cells and obtaining F.9 relation; cross-plane or cross-edition inputs cite their applicable crossing records. None of these declarations establishes performed evaluation or a verdict. |
| CC‑G4‑05 | If an acceptance clause, operator, or flow induces numeric comparison or aggregation, it cites the relevant CG‑Spec.characteristic ids and links to legality proof refs (CSLC) in the ProofLedger; if required legality grounds are missing, declare how downstream use degrades or abstains before the unsupported operation. |
| CC‑G4‑06 | Every CAL.Flow declares its result kind and the set of gating acceptance clauses; any thinning/selection‑aid policies (e.g., ε‑front selection) are explicitly policy‑bound and do not silently replace the underlying result kind. |
| CC‑G4‑07 | Every CAL.EvidenceProfile declares the provenance anchors, evidence lanes, and currentness or loss-policy pins its use needs. It cites DefaultId.GammaFoldForR_eff for B.3/C.2.2 support-model discipline and pins any numerical Γ-fold actually used. That calculation must preserve input meanings, scales, dependencies, formal/empirical distinctions, and mapping limits; absent a common model, retain separate support and a bounded synthesis. The selected receiving model determines which results a numerical loss consequence affects, including whether an R-only restriction applies. Loss assessment alone defines neither dominance nor an acceptance threshold. |
| CC‑G4‑08 | CAL.ProofLedger links each operator, flow, or clause to its required proof or justification and explicit failure behavior. A numerical Γ-fold or loss includes its receiving quantity, input scales, dependency model, derivation or calibration, assumptions, and boundary behavior; monotonicity and boundedness alone are insufficient. A missing required model cannot be repaired by F-to-R conversion or an invented default score. |
| CC‑G4‑09 | CAL publication includes RSCR tests and Worked‑Examples sufficient to detect illegality (incl. unit laundering / ordinal arithmetic), to exercise authored acceptance/flow behavior, and to validate the authored freshness envelope when it is part of admissibility; record missing tests or examples as an auditable gap. |
| CC‑G4‑10 | Each TaskMapRef = <taskMapId, taskMapEdition> resolves one immutable map edition containing the exact CALCharterRef, task, C.22 TaskSignatureRef, and edition-bearing acceptance-clause, operator, flow, and evidence-profile refs used by selection. Changing the charter, task, signature edition, or a cited component creates a new map edition. When G.4 gates are current, G.5 consumes this exact map alongside the same TaskSignatureRef; a mismatch or unresolved ref blocks that gated selector use. The map neither constructs the TaskSignature nor embeds thresholds or duplicates acceptance semantics. |
| CC‑G4‑11 | Any method/discipline specifics are placed under G.4:4.5 Extensions as GPatternExtension blocks (stable PatternScopeId, explicit governing definition, pins, and RSCR triggers); no extension introduces competing defaults or replaces G.Core invariants. |
| CC‑G4‑12 | CAL Pack@CG-Frame includes public-id continuity records for public ids: deprecations, edition bumps, and lexical-continuity notes. It exposes refresh payload pins, including editions, policies, UTS ids, and, when present, PathId and PathSliceId, sufficient for G.11 to plan RSCR without inferring semantics from prose. |
| CC‑G4‑13 | When G.4:Ext.NQD is present, CAL.NQD[] is present and is wired only via the declared subject pattern (C.18): at minimum it pins DescriptorMapRef.edition, DistanceDefRef.edition, and InsertionPolicyRef, and it treats archive/illumination summaries as report‑only unless explicitly promoted by a CAL acceptance clause/policy. |
| CC‑G4‑14 | CAL does not mint new universal types to encode “strategy/policy”. Strategy is expressed as authored flows + acceptance clauses + policy/task pins (and downstream registry/composition in G.5); any specialization is introduced only via GPatternExtension wiring blocks or cited subject patterns. |
| CC‑G4‑15 | Every runtime example keeps the reusable Method, MethodDescription, A.6.1 declaration, each precise performer’s A.13 core, independently A.15.1-admitted dated U.Work, optional later F.6 assignment-bound attribution, actual bindings and direct participants, local result, and C.2.1 result episteme distinct. Any F.6 relation uses the same obtaining A.13 assignment. A compact account may omit only an unused assignment identifier and no consumed fact; ordinary activity outside U.Work does not enter this rule. |
| CC‑G4‑16 | Each performed acceptance or decision path names its exact local result, applies the relevant result pattern, and states the exact later-use relation. Provenance stays with A.10/G.6, currentness with G.11, assurance with B.3, and decisions with C.11; no universal evaluation-result, work-result, evidence-use, or criterion-participant relation is introduced. |
G.4:8 - Common Anti-Patterns and How to Avoid Them
-
Hidden thresholds. Avoid: embedding cutoffs in CHR prose or in operator descriptions. Prefer:
CAL.AcceptanceClausewith explicit ids and pins. -
Untyped “score(x)”. Avoid: operators with implicit units and untracked legality assumptions. Prefer: explicit CHR‑typed operator signatures + cited legality checks.
-
Silent cross-sense or cross-plane reuse. Avoid: importing expressions with distinct source-local meanings, or values across ReferencePlanes or editions, without the obtaining relation and required crossing records. Prefer: cite the exact F.17 cells and F.9 relation when meanings differ, cite the applicable plane or edition crossing records, and keep each clause bounded by its stated
ClaimScopeand window. -
Acceptance as implementation detail. Avoid: acceptance embedded in tool logic. Prefer: publish acceptance as citable CAL artifacts; downstream consumes ids.
-
Exploratory telemetry treated as dominance. Avoid: letting probe/illumination telemetry quietly become a dispatch criterion. Prefer: keep it report‑only unless a policy-bound acceptance clause expressly authorizes its promotion.
-
Declaration mistaken for execution. Avoid: treating a CAL card,
TaskMap, proof-ledger row, worked example, or evidence edge as proof that an operator ran or a verdict obtained. Prefer: recover every precise performer’s A.13 core, let A.15.1 independently admit the dated Work, and add F.6 only when the receiving claim needs to establish that the Work was performed under the same obtaining assignment; recover actual direct bindings separately. Compact wording may omit only an unused assignment identifier and no consumed fact. Keep the domain-local result and any result episteme separate from both.
G.4:9 - Consequences
- CAL operator/acceptance semantics become stable, citable CAL Pack content rather than tacit code behavior.
- Legality failures are surfaced as authoring defects (RSCR‑testable) rather than run‑time surprises.
- Downstream patterns (
G.5,G.8,G.9,G.10,G.11) can reference stable ids/pins without redefining acceptance or operator semantics. - Method pluralism is supported: multiple calculi can coexist as separate operator/flow/acceptance families, wired via Extensions rather than mixed into the core kit.
G.4:10 - Rationale
CAL sits at the boundary where typed measurement becomes actionable choice. Publishing typed and testable CAL declarations reduces semantic drift and prevents “shadow legality gates” from emerging in tools or in downstream prose.
The design separates concerns:
- CHR governs measurement typing and legality guard macros,
- CG‑Spec and CN‑Spec govern the legality gate and governance card, respectively,
G.Coregoverns Part‑G invariants and trigger/default discipline,G.4governs the CAL kit: authoring objects, publication surface, and handoff manifest.
This yields modularity (one governing definition per invariant or default), auditability (pins/ids and proof refs), and extensibility (method families attach through explicit extension modules).
G.4:11 - SoTA-Echoing
These sources contribute to the CAL clauses or extensions identified below under the stated conditions. Reconsider the affected use when its source assumptions or adopted contract change.
| Source and adopted contribution | Use in CAL | Applicability limit | Return when |
|---|---|---|---|
| Angelopoulos, Bates, Fisch, Lei, and Schuster, Conformal Risk Control, ICLR 2024 — adapt bounded monotone-loss risk control only for a CAL clause whose statistical assumptions are explicit. | C3 and CC-G4-04 require loss, risk/coverage target, calibration population/window, exchangeability or declared shift treatment, and failure/abstain behavior before such a clause is published. Only an actual application under §4.4a supplies the required binding and verdict. | “Conformal”, a calibration set, or a coverage target does not make a universal acceptance guarantee, authorize deployment, or establish that evaluation occurred. | Reopen only C3, CC-G4-04, and the one worked clause/test that claims this guarantee when its assumptions or guarantee change. |
Fontaine, Togelius, Nikolaidis, and Hoover, Covariance Matrix Adaptation for the Rapid Illumination of Behavior Space, GECCO 2020 — adapt only the need to pin descriptor, distance, insertion, archive, and reporting policy when G.4:Ext.NQD is used. | C6, G.4:Ext.NQD, and CC-G4-13 keep QD method semantics in C.18 while making the CAL wiring reproducible. | Archive occupancy, coverage, QD-score, or the presence of CMA-ME wiring is not dominance, acceptance, selection, or a runtime result. | Reopen only C6, G.4:Ext.NQD, CC-G4-13, and its one NQD example/test if the adopted descriptor/archive contract changes. |
| Wang et al., Enhanced POET: Open-ended Reinforcement Learning through Unbounded Invention of Learning Challenges and their Solutions, ICML/PMLR 119 (2020) — reject as a source of G.4 core or acceptance semantics; retain it only as an exact lineage reference for optional exploration wiring under the applicable pattern. | C6 and CC-G4-11 require an exact current pattern and present-task entry and stop condition before any exploration extension is admitted; no POET-specific rule enters the CAL core. | Open-ended generation, transfer, or progress telemetry does not become a CAL acceptance rule, task authority, or selected governor by citation. | Reopen only C6, CC-G4-11, and the exact C.19/C.23 extension block if the applicable pattern explicitly adopts a changed POET-family rule set. |
Use A.18/A.19 for comparison, partial-order and selected-set rules. Distributionally robust and broader multi-objective methods are further candidates to investigate when the receiving question needs them. Revisit only the clauses, extensions and applications that depend on a changed source claim.
G.4:11.1 - CAL architecture and publication inventory
The charter fixes the pack’s scope and assumptions. Typed operators and clauses define admissible applications; flows compose them; evidence profiles and the ProofLedger expose their required grounds. The TaskMap connects these declarations to selection. Later applications bind actual inputs and produce results under §4.4a. Extensions add only the method-specific declarations needed by that use.
G.4:12 - Relations
Builds on: G.Core (and the pattern template discipline in E.8).
Uses: G.1 (CG‑Frame Card), G.2 (SoTA Synthesis Pack), G.3 (CHR Pack), G.0 (CG‑Spec legality gate), A.19 (CN‑Spec plus direct comparison and selection patterns), A.18 (CSLC), A.2.6 (U.ClaimScope), A.6.1 (declarations and actual bindings), A.13 (precise performer core), A.15.1 (independent Work admission), A.2.1 (assignment species and occurrences), and F.6 only when the receiving claim needs to establish that the Work was performed under the same obtaining assignment, C.2.1 (result epistemes), C.11 (decision results), A.10 (provenance and bounded reliance), B.3 (assurance), G.11 (currentness), and E.18 (transformation-flow structure and GateCrossing), A.21 (gate decisions from independent check results), F.9 (actual relations between source-local meanings), F.17 (SchemeSenseCells), and E.17 (multi-view publication).
Uses (via Extensions): G.6 (EvidenceGraph/Path citation; when G.4:Ext.EvidenceGraphWiring is present), C.18 (NQD), C.19 (E/E‑LOG), C.23 (SoS‑LOG).
Used by: G.5 (selector/dispatcher), G.8 (SoS‑LOG bundles), G.9 (parity), G.10 (shipping), G.11 (refresh orchestration).
Publishes to: UTS (public ids and public-id continuity records), RSCR (tests and trigger emissions), G.5 (handoff manifest), and, as cited payload, shipped packs governed by G.10.
Constrains: any run‑time LOG implementation that executes CAL operators/flows must treat CAL artifacts as citable specifications and must not re‑invent acceptance semantics.
G.4:End
G.5 - Method-Family Registry, Dispatch and Selected-Set Result Declaration
Type: General (G) Status: Stable Normativity: Normative
Intent. Help engineers and framework authors register and dispatch rival method families, and declare selector-facing sets from a selection or an already established membership basis. A set can retain alternatives for later choice or include all its members for one named use. Preserve that plurality when reporting the result. A hidden scalar winner would change its meaning.
G.5:0 - Use this when
Use G.5 to declare a selector-facing result when several already identified candidates remain available for downstream choice, or when several already identified members are all included for one named use. The members may be method candidates or other identified objects, such as framework editions; the applicable branch below preserves that distinction. The declared result states the outcome kind, members or keyed member entries, ordering status, named use when applicable, and basis pins. It does not prove that any member improved, that work occurred, that a local choice has been made, or that the result is available to an audience.
Use Shortlist or RankedShortlist for alternatives retained for later choice. Use JointUseSet only when every named member is included for one bounded use. This joint-use branch consumes exact member identities under their own rules; it does not require MethodRef, a method-family registry row, or Method classification for framework editions or other non-Method values.
When an earlier choice or another current inclusion basis has fixed the members, use G.5-6 DeclareSetResult. Supply their exact refs, named use, inclusion conditions, ordering and sufficient basis pins. Declaring the set neither registers method families nor runs G.5-3 Select. For a JointUseSet over non-Method members, it is the required branch (CC-G5.46).
For ordinary method-family dispatch, open G.5 when two or more already admitted Methods are live under grounded selector rows for the same declared task and the current question is the selector-facing set result: which candidates remain admissible, whether the admitted comparison orders them, and whether the result must be a shortlist, narrowed handoff, abstention, or escalation. If the live question is still one local choice among available options, first constitute the exact C.11 choice assertion under its predicate. Reuse existing grounded method-family rows. Create a new reusable row only when the grouping itself must recur, carry family-level policy, be versioned, or be published. Use the crossing, evidence/reliance, assurance, stable public identity and publication branches when their conditions apply.
For Method dispatch, resolve each exact A.3.1 Method and the row’s independent grouping basis under S1 (§4.2). For generator dispatch, resolve each generator under its subject pattern and the row’s grouping basis and signature under S1′. An unresolved member or grouping basis blocks its row. For actual selection, use S3 to recover the application and its bindings. Apply its Work-admission conditions when the account asserts dated Work or the receiving contract requires it. A result declaration alone establishes neither occurrence. S5 governs any separately needed result episteme, public identity and publication claim.
G.5:0.1 - What goes wrong if missed
- rival families are compared under silent comparator drift, hidden baseline changes, or unspoken crossing costs
- the selector hides its choice of a single winner even when only a partial order is admissible
- an upstream record in
C.11,C.19, orC.24leaves the selector result implicit: downstream users cannot tell which upstream choice, pool treatment, or enactment result G.5 consumes or which set it declares - a selector adopts an exploration, open-ended, or specialization policy for architectural convenience without declaring that choice
G.5:0.2 - What this buys
- one registry that keeps rival method families distinguishable and dispatchable
- one selector result form that uses the closed
SelectorOutcomeKindrules in §4.4b and the closedSetResultFamilyset when the result is set-shaped - a trace with basis pins that DRR and SCR records can address, making the selection rationale recoverable
- one declared selector outcome with its kind, applicable public label, ordering and basis pins. Give the members or keyed entries for a set, the named use for joint inclusion, the content of a handoff, or the condition for abstention or escalation.
Registry and dispatch remain the primary concern. A set declaration can communicate their outcome or identify already established members for downstream use. The declaration neither registers families nor dispatches them.
G.5:0.3 - First-minute questions
- Which exact members and grouping or inclusion basis are already established?
- Which decisions will each recipient make? Are the listed members alternatives for a later choice, or are all designated for one named use?
- Does the declared comparison justify ordering them?
- Which eligibility and evidence conditions apply? If promising a feasible combination or menu, what joint choices are allowed and what shared limits affect them?
- What result can be handed over, and what prevents a complete result?
- Does the current question additionally claim actual selection, composition, a relation between meanings, public identity or publication? Open only the applicable branch in §4.2.
G.5:0.4 - First output
State one SelectorOutcome under §4.4b: its kind, applicable members or keyed entries, ordering, named use and inclusion conditions where required, handoff or blocking content, and sufficient basis pins. Use the quick card in §4.4c. For an ordinary result over grounded rows, direct refs to the grouping, eligibility and comparison basis plus the S3 audit refs suffice; the same compact record can carry them.
A prior C.11 choice, C.19 pool-policy result, C.24 next action or another governed inclusion basis can supply the inputs. G.5 states the resulting membership or handoff. Exact framework editions retain their own identities and use G.5-6 DeclareSetResult; E.4.PFR governs their dependency or compatibility claims. Add a stable public identity only when needed. For audience availability, use E.17’s source-backed face and E.24.PUB’s publication conditions through S5.
G.5:0.5 - Minimum ordinary slice and bounded non-use
Situation. A pump-maintenance team has two already admitted A.3.1 Methods, ThresholdTrendReviewMethod-E2 and SpectralResidualReviewMethod-E1, behind the exact project-local selector rows <ThresholdTrendReview-local, R3> and <SpectralResidualReview-local, R2>. These are MethodFamilyRowRef values: each fixes its row edition, exact MethodRef[], and declared grouping basis PumpTriageCandidateGrouping-E1. The same TaskSignatureRef=PumpVibrationTriage-T1 and effective reference scheme apply to both. The task signature requires a 24-hour series input and a 30-minute review budget, and both declared Method interfaces meet those constraints. No G.4 CAL gate is current in this ordinary case, so TaskMapRef is absent. No admitted comparator justifies ordering one above the other. The live G.5 question is now how to surface that admissible set, not which pump action a decision-maker should choose.
The minimum truthful result is:
GroundedCandidateRows = [
{ methodFamilyRowRef = <ThresholdTrendReview-local, R3>,
MethodRef = [ThresholdTrendReviewMethod-E2],
groupingBasis = PumpTriageCandidateGrouping-E1 },
{ methodFamilyRowRef = <SpectralResidualReview-local, R2>,
MethodRef = [SpectralResidualReviewMethod-E1],
groupingBasis = PumpTriageCandidateGrouping-E1 }
]
SelectorOutcome(
selectorOutcomeKind = SetResultOutcome,
setResultFamily = Shortlist,
members = [<ThresholdTrendReview-local, R3>, <SpectralResidualReview-local, R2>],
ordering = unordered,
basisPins = [<ThresholdTrendReview-local, R3>,
<SpectralResidualReview-local, R2>,
PumpTriageEligibility-E1],
auditRefs = [DRR-PumpTriage-01, SCR-PumpTriage-01],
nextUse = maintenance_method_handoff
)
What changes in practice. The team stops leaving the retained pair implicit in a comparison note and stops saying “the spectral method is best.” It emits one unordered Shortlist that another receiver can cite, with the exact survivors and basis visible, while making no local-choice, actual-use, or winning-method claim. A later receiver can request one missing comparator, use the bounded handoff, or open its separately governed decision question without rewriting either Method or inventing a winner.
Near misses and non-use. Do not use G.5 merely because several names appear in one list.
- If the Method-dispatch candidates are only labels, descriptions, cards, or unresolved references, require A.3.1 and C.2.1 before dispatch.
- If the current question is one local choice among already available options, use
C.11; if it is the policy for retaining or retiring live candidate lines, useC.19; if it is enactment planning after choice, useC.24for the plan and the applicable A.15/A.6 patterns for actual Work and operation applications. - If the current object is only a composition sketch, keep the S4 template; use B.1.5 only for a qualified composite Method and A.22 only for an independently selected Structure.
- If no rival candidate set, selector result, narrowed handoff, abstain, or escalation is current, do not open
G.5. - Open F.9, A.10, B.3, stable registry or UTS identity, and E.24.PUB only for an actual crossing, relied-on evidence, assurance claim, reusable identity, or audience-availability claim respectively; their absence does not invalidate the smaller same-scheme selector result.
G.5:0.6 - Reuse a local grouping, then register it publicly when needed
The pump team in §0.5 already dispatches through R3 and R2. That unordered shortlist requires no new row. If the grouping of both Methods must recur across triage runs, the team can define:
MethodFamilyRowRef = <PumpReviewCandidates-local, R1>
MethodRef[] = [ThresholdTrendReviewMethod-E2, SpectralResidualReviewMethod-E1]
GroupingBasis = PumpTriageCandidateGrouping-E1
EligibilityBasis = PumpTriageEligibility-E1
TaskSignatureRef = PumpVibrationTriage-T1
ComparisonBasis = no admitted ordering; retain admissible alternatives unordered
The grouping criterion is “these admitted review Methods are candidates for pump vibration triage using the stated 24-hour input and 30-minute budget.” This project-local row fixes those exact members and basis; changed membership or an action-changing policy makes R2. It requires neither a UTS row nor an AssuranceProfile when this ordinary use has no assurance gate. A later real gate still consumes its required evidence and follows its unknown/fail behavior.
If the team instead intends a stable public registry entry, its proposed public continuation can be <PumpReviewCandidates, P1>, linked by an explicit naming/continuity mapping to the local R1 grouping. P1 retains the same exact Method refs and grouping criterion, adds the typed EligibilityStandardRef for the pump task and an AssuranceProfileRef stating expectations for its declared uses, and satisfies UTS naming/publication requirements under S1 and CC-G5.6. Registration is complete only when those public-contract obligations are met. For this P1 example, PumpTriageEligibilityStandard-E1 states the two Methods’ required 24-hour series and 30-minute budget, with unknown when a required input cannot be determined. PumpTriageAssuranceProfile-E1 names the evidence expectations and failure behavior for the declared triage uses; UTS-PumpReviewCandidates-P1 supplies the public name and continuity mapping to R1. P1 registration consumes these completed declarations and that UTS entry alongside the unchanged exact member/grouping basis. A missing one leaves public registration incomplete. The assurance profile supplies no B.3 result.
R1 itself completes through RegisterFamily with the six values shown above: neither AuthoringBase nor AuthoringMinimal is activated because this local grouping has no CG-Frame use. P1 adds the three public references just described and activates UTSWhenPublicIdsMinted. Now suppose a later G.5-3 Select use is governed by PumpTriage-CG-E1, whose MinimalEvidence clause requires a valid sensor calibration for the cited vibration series. Its CN/CG editions and frame pins are required. If the calibration input is missing, the evidence predicate is unknown and this example’s gate rule returns abstain; neither successful local R1 registration nor completed public P1 registration makes that selection pass.
Making local R1 available to the project audience may itself be an E.24.PUB publication occurrence when that pattern’s conditions obtain. Conversely, the intention or declaration of P1 establishes no availability. Neither branch creates a Method or an ontic family relation. Non-Method joint-use members continue through DeclareSetResult with their existing identities and inclusion basis.
G.5:1 - Problem frame
The exact CG‑Frame card from G.1 and SoTA Synthesis Pack@CG‑Frame from G.2 name the frame, EntityOfConcernRef, ReferencePlane, source rows, and rival internally coherent method families (and sometimes generator families) that may address the same declared task. If their breadth is used as a premise, cite G.2’s pack CoverageJudgementRef and its HarvestPolicy basis. Registration does not recount cards as families; a combined method/generator coverage result does not establish method-only coverage for dispatch.
At the same time, the scale and coordinate definitions from G.3 and the typed operator-argument and result declarations from G.4 specify admissible calculi and acceptance clauses. They supply grounds for formulating eligibility, assurance, and admissibility constraints. Choosing one method from the alternatives also needs a selection basis.
You need a notation‑independent way to:
- register method families and generator families as auditable, versioned entries,
- select, compose, or fall back among them at run time for a concrete task instance,
- declare stable selected-set results, including retained-alternative and all-member results, and publish stable identities to UTS when required, and
- emit RSCR‑relevant triggers and pins without inventing new “shadow specs”.
G.5:2 - Problem
How to design a general, auditable dispatcher that:
-
preserves pluralism (families from competing Traditions retain their distinct assumptions and grouping bases) while remaining dispatchable (selection is possible and explainable);
-
does not embed algorithmic dogma in the core selector kernel;
-
when expressions carry distinct F.17 source-local meanings, requires the complete crossing path—exact local senses, an obtaining F.9 Bridge, a separate bounded-use proposition, and the appropriate reliance or assurance branch—while treating pins as audit references rather than as the crossing facts;
-
produces set-valued outcomes when only partial orders are admissible or when every named member is included for one bounded use, without confusing those meanings; when exact non-Method members already have a current inclusion basis, it declares that result without routing them through method-family selection;
-
distinguishes:
- selector object set and components (registry, selector boundary, and result-declaration records),
- universal Part‑G invariants (carried by
G.Core), - method-specific and generator-specific semantics (carried only through
Extensionsblocks).
G.5:3 - Forces
-
Pluralism vs. forced totalisation. Many selection regimes are inherently partial-order; forcing a scalar winner often creates inadmissible semantics.
-
Evidence realism vs. hard gates. Eligibility and acceptance frequently depend on incomplete evidence; selection must remain auditable under tri-state unknowns.
-
Reuse vs. leakage. Reuse across distinct source-local meanings remains valuable, but it starts from exact F.17 cells and an obtaining F.9 Bridge, then keeps the proposed use, direction, rule, tolerated loss, reliance or assurance, and actual selector use separate. Bridge, CL, loss, registry, bundle, or policy pins cannot silently re-ground semantics.
-
Exploration vs. exploitation. Dispatch sometimes must probe alternatives under explicit policy envelopes and risk envelopes, but probing must not become an implicit fourth status.
-
Evolvability vs. churn. Registries evolve (new families, deprecations, edition bumps); continuity must not be broken by “rename by meaning”.
G.5:4 - Solution
G.5:4.1 - G.Core linkage (normative)
Builds on: G.Core (Part‑G core invariants; Default Governing Definition Index citation)
GCoreLinkageManifest (normative).
For the operation in use, expand the applicable profile and set ids by union with its explicit deltas (per G.Core:4.2.1). The activation conditions below select those ids before expansion; Nil‑elision does not waive an activated obligation. Select retains its exact task, row editions and DRR/SCR-addressable audit result. DeclareSetResult instead consumes its exact result family, identified members, inclusion basis, ordering and named use where required; it acquires no TaskSignature, Method or registry row merely by declaring that set.
Select profile activation before expanding the G.Core sets. RegisterFamily always retains S1’s immutable row edition, exact admitted members, grouping criterion and applicable eligibility/comparison basis. A project-local row without a CG-Frame activates neither AuthoringBase nor AuthoringMinimal; it still satisfies those S1 obligations. Intentional public registration adds EligibilityStandardRef, AssuranceProfileRef and UTS obligations even when no CG gate is in use. When a real CG-Frame registry or Select use is current, both authoring sets apply in full; omitting their required CN/CG pins is a failure, not nil-elision.
For crossing-aware selection, CorePinsRequired below lists the crossing pins individually. Each conditional pin is mandatory when its stated condition holds. When consuming G.7 calibration records or a named B.3 assurance account, retain all pins, editions, and evidence required by that account, including CC‑G7‑SCRLinkage‑1 for cited calibration evidence.
-
CoreConformanceProfileIds :=GCoreConformanceProfileId.PartG.AuthoringBase(when the current operation authors a registry in a selected CG-Frame or performs G.5-3 Select within that frame)GCoreConformanceProfileId.PartG.TriStateGuard(when evaluating eligibility or acceptance predicates)GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted(when public identities are minted or evolved, including intentional public registration under S1/S1′ and CC-G5.6; a reusable project-local row alone does not activate this profile)GCoreConformanceProfileId.PartG.ShippingBoundary(when an output is shipped)
-
CorePinSetIds :=GCorePinSetId.PartG.AuthoringMinimal(for the same CG-Frame registry-authoring or G.5-3 Select use; local registration or a set declaration alone does not activate it)
-
CorePinsRequired :=(additions to the PinSets; identify pins and refs by id only. When a use makes an optional pin mandatory, state that change. Prefer referencing inherited pins to repeating them)-
TaskSignatureRef(the C.22 TaskSignature edition forSelect; seeG.5:4.2, S2) -
TaskMapRef?(exact G.4 map edition, only when this selection uses G.4 CAL gates) -
MethodFamilyRowRef[](exact<MethodFamilyId, rowEdition>values when method-family rows are consumed or registered) -
MethodRef[](exact A.3.1 Methods resolved from every method-bearing registry row consumed or registered) -
SelectedStructureRef[]?(exact independently selected A.22 Structures consumed only when their organization changes this selector use) -
GeneratorFamilyRowRef[]?(exact<GeneratorFamilyId, rowEdition>values when generator families are in scope) -
PathId[]?,PathSliceId[]?(when audit or evidence citations use a G.6 graph, or an independently applicable gate or shipping contract requires those citations) -
UTSRowId[]?(when the operation mints, evolves or consumes a public identity; intentional public registration activates S1/S1′ and CC-G5.6 obligations) -
FailureBehaviorPolicyId?(only when degrade or abstain behavior is explicitly policy‑bound) -
SoSLogBranchId?(only when degrade or abstain behavior is explicitly policy‑bound) -
BridgeId/BridgeCardId?(the obtaining Bridge used by this selection; a Bridge Card is cited only when that Card is relied on) -
BridgeMatrixId?(when this selection uses a BridgeMatrix) -
CL/CL^k/CL^plane?(the applicable values when cited or required by the consumed calibration or named assurance account) -
Φ/Ψ/Φ_plane policy-ids?(the applicable policy ids and editions when required by the consumed calibration or named assurance account, or when crossing or plane penalties are applied) -
CrossingBundleId?(when the selector cites a CrossingBundle or its named downstream use requires one underE.18orCC‑G5.27)
-
-
DefaultsConsumed :=DefaultId.GammaFoldForR_effDefaultId.PortfolioModeDefaultId.DominanceRegime
-
RSCRTriggerSetIds :=GCoreTriggerSetId.RefreshOrchestration(payload: exact changed source and affected-use scope;SCRId,DRRId;TaskSignatureRef,CGSpecRef.editionandCNSpecRef.editionfor aSelectresult; the exactMethodFamilyRowRef[]and/orGeneratorFamilyRowRef[]consumed; and the applicableTaskMapRef?,AcceptanceClauseId[]?,SoSLogBranchId?,FailureBehaviorPolicyId?,DescriptorMapRef.edition?,DistanceDefRef.edition?,TransferRulesRef.edition?,InsertionPolicyRef?,PathId[]?,PathSliceId[]?,RSCRTestId[]?. Nongraph scope uses the existing G.CorePatternScopeIdbranch.)
G.5:4.2 - Dispatcher and Registry object set (notation‑independent)
G.5 defines the object-set components below. Their purpose is to make dispatch possible and auditable without embedding any method-family semantics in the selector kernel.
S1 — MethodFamily Registry (design-time; project-local or within a selected CG-Frame).
A reusable row represents one declared grouping. Choose its registry-identity contract: project-local reuse, or intentional registration under a stable public registry identity. This distinction concerns the identity contract, not who can see the row. Both branches fix these replayable values:
Identity and continuity:MethodFamilyIdnames the continuing row lineage;rowEditionnames one immutable edition;MethodFamilyRowRef := <MethodFamilyId, rowEdition>designates that edition. Lineage and Tradition notes andUTSRowIdremain descriptive or publication values.Exact method members: non-emptyMethodRef[], each resolving to oneU.Methodalready admitted under A.3.1.Grouping basis: exact claim, criterion, or direct relation reference that justifies this row’s grouping for the current selector use; if no ontic family or membership relation is directly governed, the basis is explicitly project-local and creates none.
One exact row edition fixes its method members, grouping basis, and every selection-changing pin. Changing any of those values creates a new rowEdition; retain the MethodFamilyId only while the declared grouping remains the same continuing row lineage. Old MethodFamilyRowRef values continue to resolve their old editions. Add task, eligibility, policy, scheme, source, ClaimScope, validity, or intended-use pins only when they change selection or a named receiver needs them; none replaces the members or grouping basis.
Eligibility and comparison basis: the actual rule and applicable editions used for this selection, including whether any comparison justifies ordering. A local row may cite its existing project rule; it need not manufacture a public eligibility artifact.Assurance expectations, only when the local use requires them, with the applicable evidence, unknown and failure rules. A real assurance or minimal-evidence gate keeps every input required by its governing clause.
Public-registry continuation. Intentional registration under a stable public identity additionally requires EligibilityStandardRef as a typed predicate record (tri-state per G.Core, using CHR/CAL terms and applicable edition pins), AssuranceProfileRef for declared evidence-lane expectations and assurance-lane pins, and the UTS naming/continuity obligations of CC-G5.6. The AssuranceProfile states expectations; it is not a B.3 assurance result. The public contract retains the same immutable members and grouping basis. The remaining fields below apply when their subject conditions hold in either branch.
AdmissibilityBindings: when the row is authored for an actual CG-Frame or consumed by its gate, cite its single governance card and gate (CNSpecRef,CGSpecRef) and every required admissibility constraint, such as scale/unit conditions for a measurement use. An ordinary local grouping with no such use adds no CN/CG reference.EvidencePins: cite what supports the asserted claim about the registered members. When applicability varies with the task, identify the supported member–task combinations and any grounds for transfer under C.23:4.2. An ordinary grouping can continue to list candidates while their applicability is being investigated. CiteG.6PathIdorPathSliceIdvalues when that support is represented in a graph or an independently applicable receiving contract requires those citations.CrossingAllowance: references to the exact F.17 endpoint senses, one obtaining F.9 Bridge, the separate C.2.1 bounded-use proposition, and the current A.10 or B.3 reliance basis, plus CL or observed-loss evidence when material, only when expressions with distinct recovered source-local meanings are actually related for this selector use. These are audit references; the field makes none of the referenced facts obtain.
For an actual crossing, first resolve both exact F.17 SchemeSenseCell endpoints and establish the two-participant F.9 Bridge under its own predicate profile. Then identify a separate C.2.1 episteme whose exact EntityOfConcern is that Bridge and whose ClaimGraph states the proposed use u, direction d, use-specific rule r, tolerated loss t, and polarity. For ordinary reliance require the matching current A.10 evidence-provenance path and local RelianceDisposition; when an actual named assurance claim is current, use B.3’s separate assurance branch. A consequential use without one retains its direct governing rule. Observed loss and CL are evidence, defeater or assurance-policy material, not Bridge participants or permission. Authorization and the actual Select application remain with their subject patterns. A Bridge id, CrossingAllowance, registry row, policy pin, CrossingBundle, DRR or SCR entry cannot substitute for any step.
PolicyHooksRef?: optional pointers to policy records (not defined here; wired via Extensions).
Here “a registry row represents a family” means that the row is the auditable selector-facing record for one declared grouping. The family id preserves that row lineage; the row ref selects one immutable edition for replay. Neither value identifies the grouped Methods, makes a membership relation obtain, or turns a common label, shared description, lineage note, eligibility rule, maturity card, evidence record, or policy into a method-family fact. Changing a row edition changes the registry artifact; it changes a Method or a separate family relation only when that object’s direct identity rule or the relation’s predicate independently says so.
S1′ — GeneratorFamily Registry (design‑time; optional; per CG‑Frame).
A reusable row groups generators of tasks and environments, which may co-evolve solver families. It uses the same local/public identity-contract distinction as S1 while retaining its own generator member and signature rules:
Identity and continuity:GeneratorFamilyIdnames the continuing row lineage;rowEditionnames one immutable edition;GeneratorFamilyRowRef := <GeneratorFamilyId, rowEdition>designates that edition.UTSRowIdis required by intentional public registration; project-local reuse does not require it.Exact generator members: non-empty references, each resolving to a generator already identified under its subject pattern.Grouping basis: the independently established classification, membership relation, or explicit project-local criterion that groups those generators for this selector.GeneratorSignatureRef: conceptual input and output semantics plus budget semantics.EnvironmentValidityRegionRef?: pinned constraints for generated environments or tasks.TransferRulesRef.edition?: required when the Open-Ended mode is enabled (semantics come from the cited extension refs).CouplerRefs?: exactMethodFamilyRowRef[]values that may be coupled with this generator-row edition.
Both branches also fix the applicable eligibility/comparison basis and every selection-changing source or policy pin; assurance expectations are conditional on the use. Changing generator members, grouping basis, or another selection-changing pin creates a new generator rowEdition; old GeneratorFamilyRowRef values continue to resolve their old editions. Intentional public registration activates CC-G5.6 naming, continuity and UTS obligations, with the applicable generator signature and use contracts. It does not import A.3.1 Method membership into generator rows.
S2 — C.22 TaskSignature input and conditional G.4 map.
C.22 constitutes the TaskSignature episteme and defines its edition rule. G.5 consumes its TaskSignatureRef and does not reconstruct it from a task, CAL pack, or map. Its edition identifies the task basis used by selection.
When this selector uses G.4 CAL gates, it also consumes one exact TaskMapRef. Resolve that immutable map edition, require its taskSignatureRef to equal the C.22 TaskSignatureRef supplied to this selector, and follow its exact CALCharterRef and edition-bearing clause, operator, flow, and evidence-profile refs. The map carries references, not a duplicate TaskSignature or clause thresholds. If no G.4 gate is current, omit TaskMapRef; an ordinary selector does not need a CAL pack merely to return a truthful bounded result.
For the G.4 safety example, G.5 receives TaskSignatureRef=SafetyPortfolioTaskSignature-E4 and TaskMapRef=<SafetySelectionMap, E3>. The map resolves CALCharterRef=<SafetyCALCharter, E2> and the cited gate declarations. A different signature ref or an unresolved charter or component blocks only that gated selector use.
S3 — Selection kernel boundary (run‑time; policy‑governed).
A notation‑independent selector that:
- consumes
TaskSignatureRef, exact method- or generator-family row refs, pinned spec refs, and an exact matchingTaskMapRefonly when G.4 CAL gates are current, - applies the declared eligibility conditions and any assurance gates required for this use (tri-state),
- computes an admissible (possibly partial) order,
- returns an outcome under §4.4b over exact candidates admitted by its method- or generator-family rules, preserving the declared ordering and
PortfolioMode(perDefaultId.PortfolioModeand explicit overrides). The kernel’sJointUseSetbranch is limited to admitted Methods; non-Method joint inclusion usesDeclareSetResultunder CC-G5.46, - emits audit records with pins addressable by DRR and SCR records.
When TaskMapRef is present, resolve its exact immutable G.4 map edition before applying any cited gate. Its taskSignatureRef must match this selector’s C.22 TaskSignatureRef; its CALCharterRef must recover the CG frame, EntityOfConcern, ReferencePlane, specification editions, and assumption envelope; and each cited clause, operator, flow, and evidence profile must resolve at its exact edition. Carry the exact map ref among the result basis and refresh pins. Do not copy thresholds or acceptance semantics into G.5.
If a cited gate consumes R, resolve the quantity and support model under CC‑G5.4 before evaluating its threshold. Keep formal and empirical inputs, required premises, complementary support, dependence, scope, and counterevidence distinguishable. No common model means no invented aggregate: retain a qualified synthesis, and apply the clause’s unknown behavior only where that missing quantity is actually required. The ordinary selector result in §0.4 is not an assurance claim and needs no new R calculation.
For every MethodFamilyRowRef consumed here, resolve the exact immutable row edition and then its A.3.1 MethodRef[], grouping basis, and selection-changing pins before admitting the candidate. For every GeneratorFamilyRowRef, resolve its immutable edition, generator refs under their subject patterns, grouping basis, signature, and selection-changing pins under S1′. The selector may compare or return exact row refs as auditable selector-facing addresses, but row selection neither creates its members nor proves that every listed member belongs, is admissible, is selected, or will be enacted. An unresolved member reference or missing grouping basis blocks that row’s use; it is not repaired by a label, description, UTS identity, policy, or evidence pin.
When a selector consumes an organization among Methods, cite an exact SelectedStructureRef only after A.22 has independently identified the U.Structure from exact constituents, exact already-obtaining relation occurrences, applied constraints, and one named use frame. G.5 neither supplies those discriminators nor selects the Structure by listing it. If the organization instead constitutes one composite Method, consume the exact A.3.1 Method only after B.1.5 has qualified that candidate from its independent parts and whole-forming basis.
S3 states reusable selector behavior. A completed admitted selection establishes the A.6.1 Select application, its effective argument bindings and the A.19 SelectionSlot binding for the set returned by value. The eligibility evaluation has its own application and guard result; on abstain no Select act proceeds. Keep the actual eligibility and comparison grounds recoverable.
When the account also asserts a dated selector U.Work, or a receiving contract requires that identity, first recover each precise performer’s A.13 core for the action, scope, working situation and window, including the obtaining assignment used by any exact attribution. A.15.1 independently admits the Work from its performance history, enacted Method, temporal extent and containing-System relation. F.6 applies afterward only for a required exact assignment-bound attribution through the same obtaining assignment.
A compact Work account may omit an assignment identifier unused by its receiving claim; it retains every criterion, classification, assignment, admission or attribution fact the claim consumes. Holder equality or overlapping times alone does not establish the F.6 relation.
Replay upstream CPM applications through their own declaration-local bindings. Separately admitted comparison or selection Work is required when that Work is asserted or consumed. Evidence use, A.10 reliance/provenance, G.11 currentness and a C.2.1 result episteme retain their independent grounds when asserted or consumed by the receiving use.
S3.A — TaskFamilySpecializationProfile@Context (run‑time; conditional).
When the selection question is whether or how candidates can acquire usable specialization on a declared task family, the selector may emit one TaskFamilySpecializationProfile@Context for each candidate, one SpecialistHandoff, or one narrowed handoff plan. Here profile means one selector-time comparison record for bounded specialization, not a new U-kind and not a generic narrative profile. Use the adaptation-signature fields defined by C.22.1 for this comparison. Identify who performed the operation and which help was available in the check and will be available in the receiving work. Carry an observed change on a second task at the strength warranted by the transfer evidence.
The profile should therefore cite one AdaptationSignatureRef or equivalent pinned field set carrying the declared TaskFamilyRef or TaskSignature, the work-measure threshold target, prior exposure declaration, time-to-threshold, budget-to-threshold, post-threshold efficiency when relevant, any declared transfer or retention claim, any downside cost or downside on adjacent tasks, and any specialization-entry baseline, specialization-entry evidence, or stepping-stone evidence item that materially affects comparison.
Admission rule for SpecialistHandoff: use that handoff kind only when the truthful declared result is one heterogeneous handoff bundle whose members occupy different specialization positions that still need to travel together. Do not use it when a SetResultOutcome with Shortlist, RankedShortlist, or JointUseSet, or a HandoffOutcome with another admitted handoff kind, already states the result more precisely.
When the declared task family is heterogeneous, the selector may return one SpecialistHandoff, one other narrowed handoff plan, or one SetResultOutcome with an admitted SetResultFamily that preserves rival specialists rather than declaring a single winner without grounds. Low-human-overlap candidates remain admissible only when the profile, evidence basis, and policy constraints are explicit.
S4 — Composition and fallbacks templates (design‑time).
A library of composition shapes—preconditioner -> solver -> verifier, cascades, and meta-selectors—remains available as design-time templates, admissibility-checked and pinned. A template is a description or policy-bound arrangement for possible composition; its existence, diagram order, registry placement, or selection does not create a Method, methodPartOf occurrence, obtaining relation, or selected Structure.
If exact A.3.1 Methods, exact B.1.5 methodPartOf occurrences, all other required whole-forming claims and constraints, whole semantics, interface boundary, and reidentification rule qualify one already identified candidate as a composite U.Method, consume that exact Method through the B.1.5 branch. If independently identified Methods and already-obtaining relations are instead organized for one use without constituting one Method, consume an independently selected A.22 U.Structure only after its exact constituents, selected obtaining relation occurrences, applied constraints, and named selection-use frame are present. MethodRelationStructure may remain a local readable designator for that actually selected Structure; it is not a U-kind, relation kind, Method holon, registry-row identity, or generic @BoundedContext object.
A C.2.1 episteme may describe either governed object. A.3.2 applies only when the episteme’s EntityOfConcern is one already admitted Method and its claims substantively describe that Method; an episteme whose concern is the selected Structure is not thereby a U.MethodDescription. Concrete strategy semantics stay in the referenced method families; G.5 only carries the composition template, selector relation, registry row, exact consumed Method or Structure reference, or selected-set result. None of those G.5 artifacts supplies the B.1.5 or A.22 construction facts.
Use C.29 when a selector representation or its transfer leaves an unresolved correspondence, preserved-or-lost distinction, or inference needed for the current question. An adequate local calculation or representation may be used directly.
S5 — Result, public identity, and telemetry record boundary (run-time).
Declare the following S5 outputs:
DRR(decision rationale) andSCR(evidence and confidence citation) with explicit pins,- declared selector and selected-set records produced either by
G.5-3 Selectover grounded method- or generator-family rows or by the already-grounded-memberG.5-6 DeclareSetResultbranch, - telemetry pins to refresh orchestration (
G.11), without governing orchestration.
S5 governs the selector-facing record boundary, not truth or actuality by record existence. A DRR, SCR, selected-set record, shortlist id, telemetry event, refresh cue, policy pin, or result label does not create dated Work, an actual operation application, the selected-set binding, a domain result, an evidence-provenance relation, assurance, authorization, or publication availability. Persist a selector-result claim as its own C.2.1 episteme when another use must rely on it; connect evidence through A.10, assurance through B.3, authorization through its direct governor, and actual availability through E.24.PUB only when each claim has its independently established basis.
Use §4.4b for outcome kinds and §4.4c for conditional public-identity fields.
S6 — Governance and evolution declaration boundary (design-time).
State how registry entries are versioned, deprecated and published through UTS, and how continuity is maintained as entries change. These changes use the existing Part-G types.
G.5:4.3 - Selector head and narrower selector families
Selection and dispatch stay one generic selector head. Narrower selector families may refine it, but they do not redefine the universal invariants pinned through G.Core, do not add new mandatory inputs to inherited Select, and do not mutate inherited SlotKinds. Required policy and edition refs use the declared input meanings.
Method- and generator-specific pressures such as QD archives, open-ended declared sets, explore and exploit lenses, or preference comparators do not become part of the selector head. They arrive only through explicit extension declarations and the pins those extensions require.
G.5:4.4 - Selector Relation Fields
| Selector relation | Consumes | Produces |
|---|---|---|
| G.5-1 RegisterFamily | declared local or public registry-identity contract; continuing MethodFamilyId and new immutable row edition; nonempty exact admitted A.3.1 MethodRef[]; obtaining grouping relation or explicit grouping criterion; eligibility/comparison basis; selection-changing source, policy and CHR/CAL/CN/CG pins as applicable; public-continuation fields when selected | One immutable MethodFamilyRowRef = <MethodFamilyId, rowEdition>, resolving the members, grouping, eligibility/comparison basis and applicable pins. Public registration additionally fixes EligibilityStandardRef, AssuranceProfileRef and UTSRowId under S1/CC-G5.6. A local row retains conditional assurance expectations without requiring a public UTS entry. Neither branch creates Methods or grouping facts. |
| G.5-2 RegisterGeneratorFamily | declared local or public registry-identity contract; continuing GeneratorFamilyId and immutable row edition; nonempty exact generator refs under their subject patterns; grouping basis; GeneratorSignatureRef; applicable eligibility/comparison, source and policy pins, including TransferRulesRef.edition when required | One immutable GeneratorFamilyRowRef = <GeneratorFamilyId, rowEdition>, resolving those members, basis, signature and applicable pins. Intentional public registration additionally meets S1′/CC-G5.6 naming, continuity and UTS requirements. Local reuse retains the same replayable member/basis core; neither branch creates generator identity or membership. |
| G.5-3 Select | TaskSignatureRef; exact matching TaskMapRef when G.4 CAL gates are current; exact MethodFamilyRowRef[] and/or GeneratorFamilyRowRef[] for the candidate rows in scope; each immutable method row resolves non-empty exact A.3.1 MethodRef[] under S1, and each generator row resolves non-empty exact generator refs under their subject patterns and its signature under S1′; every row resolves its grouping basis and selection-changing pins; pinned CNSpecRef and CGSpecRef editions; policy refs if any; sufficient audit basis refs, with PathId or PathSliceId only for actual graph citations or an independently applicable gate or shipping contract | For admitted selection: the actual CandidateSet return, its result binding and the declared selector outcome, with PortfolioMode, exact row/map refs and sufficient DRR/SCR basis pins. For abstention: AbstainOutcome with the failed or unknown condition and no SelectionSlot binding. An actual empty-set return retains its criteria and scope; a justified further inquiry or action may supply an ActionHint or EscalationOutcome. |
| G.5-4 Compose | CandidateSet, composition template refs, pinned admissibility constraints | Composite strategy template (template-level; admissibility-checked; pinned) |
| G.5-5 Telemetry | run outcomes, citations, and policy or edition pins | refresh cues (typed RSCR causes and payload pins), parity deltas (if parity harness is in use), telemetry pins (selector-side; orchestration governing definition is G.11) |
| G.5-6 DeclareSetResult | one exact SetResultFamily; exact already identified memberRef[]; namedUse for JointUseSet; ordering; inclusion or selection conditions; and sufficient basisPins to the already current choice, pool treatment, accepted decision, or other governed inclusion basis | one SelectorOutcome with SelectorOutcomeKind = SetResultOutcome and the exact membership form required by that family. For JointUseSet, it emits keyed unique memberEntries, ordering = unordered, the named use, inclusion conditions, and basis pins without a method-family row or Select pass. |
RegisterFamily and RegisterGeneratorFamily produce only the local or public registry row selected under S1 or S1′, respectively. They create neither member identities nor independently governed membership facts. Select may address candidates through those rows only after their exact members resolve under the applicable method- or generator-family rule and their grouping bases resolve; its returned candidate or selected-set value does not retroactively ground a row member.
Compose produces only the pinned template named in its output column. It neither qualifies one composite Method under B.1.5 nor selects one A.22 Structure. When a later selector use consumes either governed object, the exact Method or Structure reference is an independently grounded input rather than a result inferred from this template.
DeclareSetResult begins only after its exact members and inclusion or selection basis are current. An upstream C.11 ChoiceResult, C.19 pool treatment, accepted decision, or another governed basis may appear among basisPins; the G.5 branch does not repeat or perform that decision. It declares the selector-facing set-result content and stops. It creates no member identity or relation, method-family row, Select application, dated selection Work, persisted C.2.1 result episteme, assurance or authority claim, or E.24.PUB availability occurrence.
G.5:4.4a - Worked selector slice
-
A catalyst-search team is choosing among three method families for the same declared
TaskSignatureandC.22.1adaptation signature. -
For each of the first two families the team tests a research setup using the member it would choose for each task. The setups start below the common threshold, with comparable prior exposure, the same check schedule and budget, and the same assistance available in testing and in the intended work. By the first scheduled check, the first setup passes and the second still falls short. The second passes at a later check. On the adjacent tasks the second setup performs better under the same test conditions.
-
The project values both passing the early check and performance on those adjacent tasks. Each of the two eligible alternatives has an established advantage on one of these criteria. The selector returns them as an unordered
Shortlistand records the trade-off in the DRR and SCR. For the third family the evidence falls short ofMinimalEvidenceand the selector abstains. To credit the original training for performance on adjacent tasks, use the transfer comparison in C.22.1:8. -
If the project also needs to identify the dated selection Work and its precise performer, A.13 first recovers
CatalystSelectorSystem-17 : U.Systemfor exact actionCatalystFamilySelectionAction-17.CatalystSelectorBoundary-17contains the deployed selector runtime, its effective policy state, and its registry/evidence interfaces; it excludes the method-family rows,TaskMap, result records, assignment, and containing team System. The action applies the effective selector to the three candidate families and returns the retained set. Its scope isCatalystFamilySelectionClaimScope-17, its working situation isCatalystSearchSelectionSituation-17, and its window is2026-07-30T10:00:00Zthrough2026-07-30T10:08:00Z.CatalystSelectionAdmissibilityNorm-17directs the selector to exclude candidates that failMinimalEvidence, preserve admissible non-dominated alternatives, and abstain rather than manufacture a scalar winner. Relevant conditions include the exactCatalystTaskSignature-17, current row and map editions, eligibility evidence, comparison policy, and adaptation-signature values. -
A.2 declares local agential kind
CatalystMethodSelectorSystemRole. Its membership criterion requires the stable work-facing contribution of method-family selection and goal-directed, condition-sensitive regulation underCatalystSelectionAdmissibilityNorm-17: the holder must apply the current gates, preserve the admissible set-return semantics, and abstain or escalate when no candidate qualifies.CatalystSelectorDecisionTrace-17showsCatalystSelectorSystem-17excluding the third family for failedMinimalEvidence, retaining the first two as non-dominated, and emitting no scalar winner. The trace and boundary/runtime records support the criterion facts under A.2’s membership rule; A.10 makes that source-to-use account recoverable. The case independently classifiesCatalystSelectorSystem-17underCatalystMethodSelectorSystemRole; neither the assignment nor the candidate Work supplies the classification. No Grade, autonomy result, characteristic profile, or stronger assurance claim is consumed. -
The same A.13 core uses
CatalystSelectorAssignment, a directly declared species underU.SystemRoleAssignment. The species declares holder, assigned-kind, and task-signature participant meanings and the assignment predicate.CatalystSelectorAssignment-17obtains withCatalystSelectorSystem-17,CatalystMethodSelectorSystemRole, andCatalystTaskSignature-17as its exact participant values; its maximal uninterrupted predicate-true interval covers the stated scope, situation, and window. -
Only after that core is established does A.15.1 independently admit
CatalystSelectionWork-17 : U.Workfrom the exact selection-action history, enactedCatalystFamilySelectionMethod, temporal extent, and obtaining containing-System relation to independently admittedCatalystSearchTeamSystem. Actual applicationCatalystSelectApplication-17separately carries its effective candidate, criteria, and A.19SelectionSlotbindings. Neither the assignment nor F.6 is an A.15.1 admission premise. -
Because this account states that the Work was performed under
CatalystSelectorAssignment-17, F.6 afterward establishesperformedUnderAssignment(CatalystSelectionWork-17, CatalystSelectorAssignment-17)through that same obtaining A.13 assignment. The direct case fact links the exact pair, holder equality holds, and the assignment interval covers the Work. A different overlapping assignment held by the same System would not establish this attribution. A short result may omit the assignment identifier only after every fact consumed by the attribution remains recoverable. -
A persisted shortlist assertion is a separate C.2.1 episteme; its DRR or SCR references do not by themselves prove the exclusion facts, warrant the result, authorize downstream action, or make that episteme available to an audience.
-
When one upstream
C.19pass has already narrowed the live pool to one internal retained subset over registered families,G.5-6 DeclareSetResultmay declare that result as oneShortlistwith oneShortlistIdand explicit basis pins only when selector-facing result declaration is now the question. Until that declaration occurs, the internal retained subset is not yet one G.5 shortlist result. -
When one upstream
C.11pass has already fixed one local choice over one declared source set,C.19has fixed one retained pool treatment, an accepted decision has fixed all-member inclusion, orC.24has produced one enactment-facing narrowed handoff, useG.5-6 DeclareSetResultwhen selector-facing set-result content is now the question. Until that declaration occurs, theChoiceResult,PoolPolicyResult, accepted inclusion basis,CallPlan, orCheckpointReturnis not itself that G.5 result. AJointUseSetover non-Method members usesDeclareSetResultand does not pass throughRegisterFamilyorG.5-3 Select; ordinary generator selection retains the S1′/S3 branch.
G.5:4.4b - Declared selected-set result and closure rule
State the selector-facing outcome and the information its kind requires, so the recipient can use it without reconstructing it from a selector trace, comparison note, or local choice.
For method- or generator-family dispatch, that result closes selector work over grounded rows. For a JointUseSet, it records already identified members that are all included for one named use. It does not replace registry maintenance, comparison rules, the upstream choice or inclusion basis, or the patterns that identify the members and their relations.
The admissible selector outcome families here are:
SelectorOutcomeKind = SetResultOutcome, whose closedSetResultFamilyvalue set isShortlistwhen alternatives are retained for later choice and the result does not order them,RankedShortlistwhen the result orders those retained alternatives, andJointUseSetwhen every named member is included for one named use;SelectorOutcomeKind = HandoffOutcome, withHandoffKind = SpecialistHandoffor one other narrowed handoff plan when heterogeneity is the truthful downstream result;SelectorOutcomeKind = AbstainOutcomewhen required grounds are missing or no candidate is admissible and selection cannot proceed; andSelectorOutcomeKind = EscalationOutcomewhen a blocked selection warrants a specified further inquiry, decision or action by another competent receiver.
G.5-3 Select uses grounded candidates and the applicable eligibility/comparison basis. An admitted Select may return an empty candidate set under its declared criteria; a corresponding empty Shortlist retains that actual return, criteria and scope. An abstention instead records the failed or unknown condition and has no Select result binding. G.5-6 DeclareSetResult emits SetResultOutcome from exact already identified members and a current inclusion basis; it does not perform that upstream choice.
When the recipient asks which members were agreed for joint use, the agreement can settle their membership. To recommend a feasible combination, establish that the members meet their uses together under the shared constraints. If several recipients may choose independently from menus, check the joint choices their rights permit. A promise to honor them all needs one jointly feasible way for the participants to act. Each participant must act alike in situations they cannot yet distinguish when they commit, within their own authority (C.11:4.2 and C.43:4.1.1). For a claim that the recipients actually used the members, establish the relevant Work.
A JointUseSet uses this bounded representation:
namedUsestates the one joint use;memberEntriescontains one keyed entry per included member;- every entry has one exact
memberRef; the membership result adds no per-member contribution or basis field; - each exact
memberRefoccurs at most once, and entry order has no semantic effect; - if a serialization also emits top-level
members, it is only the unique set projection ofmemberRefvalues frommemberEntries, never a second maintained list; ordering, inclusion conditions, and sufficient top-levelbasisPinsremain explicit; and- candidate-pool membership and excluded candidates stay separate from emitted joint-use membership.
Exact content, claims about a member’s use or contribution, and direct relations keep their own governed records. When one supports the membership result, cite that existing record among basisPins; memberEntries creates neither the cited content nor a new contribution relation.
For framework use, memberRef may name an exact already identified edition under its existing identity rules. Do not populate MethodRef, create a registry row, or classify that edition as a Method merely to emit the result.
Every outcome states its SelectorOutcomeKind, public result kind when applicable, members, keyed entries, handoff content, or blocking condition, ordering, and sufficient basis pins. A handoff also states its next downstream use boundary.
A compact retained-alternative result may look like:
SelectorOutcome(
selectorOutcomeKind = SetResultOutcome,
setResultFamily = Shortlist,
members = [family_A, family_C],
shortlistId = shortlist_17,
ordering = unordered,
basisPins = [pathSlice_41, scr_22],
nextUse = downstream_comparison
)
A compact joint-use result may look like:
SelectorOutcome(
selectorOutcomeKind = SetResultOutcome,
setResultFamily = JointUseSet,
namedUse = cohort_review,
memberEntries = [
{ memberRef = FPF@C },
{ memberRef = Domain@D },
{ memberRef = Local@L }
],
ordering = unordered,
inclusionConditions = [all_three_editions_required_for_cohort_review],
basisPins = [choice_result_12, edition_basis_7]
)
Close with Shortlist or RankedShortlist when the result retains alternatives. Close with JointUseSet only when every member is included for the named use and that membership is established for every entry key. Close with a handoff, abstain, or escalation outcome when that is the actual result. A complete G.5 result states its basis and ordering status and supplies the fields its kind requires. For a set, these include its family, retained members or keyed joint-use entries, and the named use for joint inclusion. For a handoff, abstention or escalation, give the corresponding handoff content or blocking condition.
G.5:4.4bb - Public labels over archive, front, and style source sets
When a selector consumes a declared ExplorationArchive, Archive, Front, or Q-front, keep that object as a source-set family or source-set reference; it is not the emitted G.5 outcome. The emitted result states one admitted SelectorOutcomeKind and, for a set result, one admitted SetResultFamily. StyleShortlist and TraditionShortlist may be public domain labels over an admitted set-result family after their local meaning and any required term Bridge are established; they do not extend either closed set.
SelectedSetResultLabelLine@Context:
selectorOutcomeKind:
setResultFamily?:
sourceSetFamily?:
sourceSetComposition?:
publicSelectedSetLabel?:
namedUse?:
memberEntries?:
membersOrHandoff?:
derivedViewKind?:
basePaletteOrArchiveRef?:
ordering:
basisPins:
nextUse:
Earlier records may keep membersOrHandoff. Read it as members for Shortlist or RankedShortlist and as handoffContent for a HandoffOutcome. It cannot replace keyed memberEntries in a JointUseSet; if it also lists joint-use members for compatibility, that list is only the unique set projection of the entry keys.
Use sourceSetFamily for a single immediate source and sourceSetComposition for a composition, under §4.4e; retain a primary family when the receiving use distinguishes one.
sourceSetFamily may name a declared Front, Q-front, ExplorationArchive, Archive, current pool subset, or derived tradition view. For retained alternatives, publicSelectedSetLabel normally names Shortlist or RankedShortlist and may use a domain label such as StyleShortlist or TraditionShortlist only when its local meaning and any required term Bridge are established. JointUseSet is not a shortlist label: it names an all-member result and therefore uses namedUse plus keyed memberEntries. G.5 does not create the archive, compute the comparison, govern the pool policy, decide the cultural-evolution case, establish member identity or relations, or repair the term bridge. Use C.18 for archive formation, A.19.CPM for comparison, C.19 for pool policy, C.36 for cultural-evolution claims, each member’s own identity and relation patterns for those facts, and F.17/F.18/F.9 for local meanings, naming settlement, and any obtaining term Bridge.
G.5:4.4c - Result-declaration quick card
Use the outcome definitions in §4.4b and fill only the applicable fields:
| Field | When and what to state |
|---|---|
selectorOutcomeKind | Every result: the admitted set, handoff, abstain or escalation kind. |
setResultFamily, members | For retained alternatives: the admitted shortlist family and exact surviving refs; preserve a justified order for RankedShortlist. |
setResultFamily, namedUse, memberEntries, inclusionConditions | For joint inclusion: JointUseSet and its §4.4b keyed membership declaration. |
handoffKind, handoffContent | For a handoff: SpecialistHandoff or another admitted narrowed handoff and the content the next receiver needs. |
blockingPins | For abstain or escalation: the actual blocking conditions. |
ordering | Ranked, unordered or not applicable, as the outcome permits. |
basisPins, nextUse | The supporting basis and next use boundary; none when there is no next use. |
publicId | Only when stable public identity is needed; ShortlistId is specific to a shortlist. |
The pump result in §0.5 and the joint-use declaration in §4.4b show complete filled forms. A missing required value leaves the result incomplete.
G.5:4.4ca - Derived tradition-view result stays derived over one declared palette
When the source is TraditionFront or TraditionArchive, keep its base SoTAPaletteDescription recoverable. State SourceSetFamily; add DerivedViewKind when it changes interpretation or later publication and SourceSetComposition only when several declared source sets were actually composed. Cite the derivation’s declared Q, reachability or coverage rule among the DRR/SCR or equivalent basis pins. The view qualifies the source; §4.4b still defines the emitted outcome.
G.5:4.4d - Worked result-declaration closure slice
| Receiving situation | Complete result and changed action |
|---|---|
| The two pump Methods in §0.5 survive, with no admitted ordering. | Emit its unordered Shortlist; the receiver still has a choice to make. |
| A declared comparator orders family_B before family_A for the specialist handoff. | Emit a RankedShortlist with [family_B, family_A], the comparator and supporting basis pins, and the handoff use. A request for an order alone supplies no comparator. |
The cohort decision includes FPF@C, Domain@D and Local@L together. | Emit the §4.4b JointUseSet with all three editions and the conditions under which the cohort is to use them. |
| Required selection grounds are missing, or the established gates exclude every candidate. | Emit AbstainOutcome with the respective unknown or failed condition and its basis. Use EscalationOutcome only for a justified further inquiry or action. If an admitted Select application instead returns an empty set under its declared criteria, preserve that actual result and explain its scope. |
The following extensions apply only when their corresponding mode is active. Their declared Uses and pins cite the governing semantics.
GPatternExtension block: G.5:Ext.EELog
-
PatternScopeId:G.5:Ext.EELog -
GPatternExtensionId:EELog -
GPatternExtensionKind:MethodSpecific -
GoverningPatternId:C.19 -
Uses:{C.19} -
⊑and⊑⁺:∅ -
Required pins, edition pins, and policy pins (minimum):
EELensPolicyRef(or equivalent lens or policy id carried byC.19)RiskBudgetRef?ProbeAccountingRef?FailureBehaviorPolicyId?(if degrade behavior is governed by policy)
-
RSCRTriggerKindIds:{RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent} -
Notes (extension discipline; semantics cited):- Use this block for dispatch under an exploration and exploitation policy.
- Examples include bandit-style or Bayesian selection under explicit risk budgets, adaptive probing, and exploration with policy-bound abstain or degrade behavior.
GPatternExtension block: G.5:Ext.SoSLOG
-
PatternScopeId:G.5:Ext.SoSLOG -
GPatternExtensionId:SoSLOG -
GPatternExtensionKind:MethodSpecific -
GoverningPatternId:C.23 -
Uses:{C.23} -
⊑and⊑⁺:∅ -
Required pins, edition pins, and policy pins (minimum):
SoSLogRuleId[]SoSLogBranchId[](including escalation branches, if used)FailureBehaviorPolicyId(if degrade behavior is made explicit)MaturityRungId[]?(when maturity ladders are used as gates; semantics come fromC.23)AdmissibilityLedgerRef?(when selector consumes admissibility rows rather than recomputing thresholds)
-
RSCRTriggerKindIds:{RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.MaturityRungChange, RSCRTriggerKindId.EvidenceSurfaceEdit} -
Notes (extension discipline; semantics cited):- The pinned ids identify the rule and branch that governed each dispatch decision. Using a branch leaves the three acceptance statuses unchanged.
GPatternExtension block: G.5:Ext.NQD
-
PatternScopeId:G.5:Ext.NQD -
GPatternExtensionId:NQD -
GPatternExtensionKind:MethodSpecific -
GoverningPatternId:C.18 -
Uses:{C.18, C.19} -
⊑and⊑⁺:∅ -
Required pins, edition pins, and policy pins (minimum):
DescriptorMapRef.editionDistanceDefRef.editionInsertionPolicyRefTaskSignatureRef(when QD is enabled via TaskSignature flags or traits)- active fields from C.21’s DHC replay basis (only when this telemetry consumes a C.21 DHC coordinate; carry exactly the fields that coordinate used)
-
RSCRTriggerKindIds:{RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent} -
Notes (extension discipline; semantics cited):- G.5 core remains QD‑agnostic; QD semantics are governed by
C.18. - Examples include MAP-Elites-class QD, CMA-ME-class hybrids, and illumination or coverage telemetry whose descriptor and archive policies affect replay.
- G.5 core remains QD‑agnostic; QD semantics are governed by
GPatternExtension block: G.5:Ext.OpenEndedFamilyWiring
-
PatternScopeId:G.5:Ext.OpenEndedFamilyWiring -
GPatternExtensionId:OpenEndedFamilyWiring -
GPatternExtensionKind:GeneratorSpecific -
GoverningPatternId:G.2 -
Uses:{G.2, C.19, C.23} -
⊑and⊑⁺:∅ -
Required pins, edition pins, and policy pins (minimum):
GeneratorFamilyRowRef[]TransferRulesRef.edition(mandatory when Open‑Ended is enabled)EnvironmentValidityRegionRef?CoEvoCouplerRef[]?SoSLogBranchId[]?(when validity of generated tasks is gated by explicit branches)
-
RSCRTriggerKindIds:{RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent} -
Notes (extension discipline; semantics cited):- This block enables declared sets of
{Environment, MethodFamily}pairs without redefining generator semantics in G.5. - Examples referenced through G.2 family cards include POET-class and other open-ended or co-evolutionary regimes whose transfer policies and validity gates must be edition-pinned.
- This block enables declared sets of
G.5:4.4e - Source sets, operating modes and comparison policy
Use §4.4b for the emitted outcome and §4.4bb–ca for its source and public-label interpretation. An actual SelectionSlot binding carries the by-value selected candidate set; it is separate from G.5’s declared SelectorOutcome. ChoiceSet remains an ordinary mathematical set gloss, not an additional public result kind.
| Declaration | Meaning and applicable condition |
|---|---|
Front | The non-dominated source set under the declared DominanceSet. |
Archive | The exploration set retained under its policy. |
PortfolioMode | How the selector operated. The default Archive retains exploration evidence; it establishes neither an emitted Archive nor a different result family or DominanceSet. |
SourceSetFamily, SourceSetComposition | State the family of a single immediate source. For a composition, identify each contributing source and family and how its members contribute, for example a front and an archive used through their union. Retain a primary source and its family only when the receiving use distinguishes one, under A.19.SOURCE-SET-SPACE-SUBSTRATE:4.7. |
DerivedViewKind, BasePaletteRef | Qualify an actual derived view under §4.4ca; the latter is a reference, not a kind. |
PromotionPolicy | Required when tie-break or telemetry signals are promoted into dominance. |
SubjectKind, RetentionIntent=steppingStone | Qualify the relevant declaration or retention policy; neither names another emitted set result. |
Use controlled tokens, cited ids or already declared head labels for these fields. CostToProbe, ValueOfInformation, ValueOfComputation, explore_share, graduation conditions and sequencing pressure belong to the surrounding choice doctrine when they affect the decision; a result field does not establish them. All-member membership and candidate/exclusion records retain the separation in §4.4b.
G.5:4.6a - Causal method dispatch declarations
When method dispatch compares causal uses, each compared Method declares its causal question/rung and whether it is being used as an observational predictor, intervention optimizer, counterfactual strategy, causal fairness estimator, causal-RL policy, or simulation-only Method.
MethodFamily.causalUseDispatchSpec?:
causalUseQuestionRef?: CausalUseQuestionRef
targetCausalityLadderRung: CausalityLadderRung
causalUseClaimKind: CausalUseClaimKind
causalActionPolicyClass?: CausalActionPolicyClass
causalSupportComponentRefs?: CausalSupportComponentRefs
causalUseSupportResultRef?: CausalUseSupportResultRef
causalMethodUseClassification:
observationalPredictor |
interventionOptimizer |
counterfactualStrategy |
causalFairnessEstimator |
causalRLPolicy |
simulationOnlyMethod
supportedUse
unsupportedUse
CausalUseQuestionRef identifies the question content used by C.28; it is not a durable root U-kind. causalMethodUseClassification describes the Method’s proposed selector-facing use and supplies no system-role assignment, responsibility, authority, or causal certification.
A simulation-only Method cites simulationResultRef inside its support components and states bounded model use plus unsupported realized/interventional use. G.5 declares the dispatch result; C.28 supplies the causal-support result. A selector may still abstain even when a C.28 result is supported.
G.5:5 - Archetypal Grounding
Tell (archetype). The selector-bearing System selects among rival families under the declared measurement, crossing and evidence conditions. When a result Episteme is needed, it keeps the comparison basis, audit pins and refresh conditions recoverable. A joint-use declaration states that all its named members are included for its stated use. Describing them as alternatives would leave any one of them optional.
Show 1 (multi-Tradition dispatch; unordered shortlist).
A CG-Frame includes multiple decision-theoretic families with different admissibility assumptions. Evidence for some CHR traits is incomplete.
The selector-bearing System registers the families (S1), then runs Select (S3) on a pinned TaskSignatureRef. Eligibility is tri-state; missing minimal-evidence pins require abstention for some families. Among remaining candidates, only a partial order is admissible, so the selector emits one Shortlist with explicit basisPins instead of inventing one scalar winner. No shadow acceptance logic appears in the selector; it consumes pinned acceptance and admissibility records.
Show 2 (specialist handoff; ranked result).
A bounded-specialization comparison keeps two method families live under a pinned admissible comparator that orders them, and downstream handoff needs that ordering.
Declare one RankedShortlist with that ordering and comparator among its basis pins, ShortlistId when public identity is needed, and handoff-facing nextUse. If no admissible comparator supplies an order, retain an unordered Shortlist; the request for a ranked handoff does not establish one.
Show 3 (no admissible survivor; abstain or escalation). In this frame, one admissibility gate and one minimal-evidence gate fail at the same time. The failed gates block selection. Report the blocking pins and the next downstream-use boundary. Abstain, or escalate when a further inquiry or action is justified. An unexplained empty shortlist would hide the difference between this stop and an admitted selection that returned an empty set under its criteria.
Show 4 (complementary framework editions; unordered joint use).
A training cohort needs FPF@C, Domain@D, and Local@L together. The editions are already identified under their own edition rules; they are not Method candidates or registry rows. An accepted cohort decision supplies the exact members and basis. G.5-6 DeclareSetResult emits one unordered JointUseSet with one keyed entry per edition, the named cohort-review use, inclusion conditions, and sufficient top-level basis pins. Direct dependencies and pairwise compatibility claims remain with E.4.PFR; publication and access remain with E.17/E.24.PUB and the applicable access-carrier pattern. The G.5 result declares membership but does not perform the choice, make those neighboring claims obtain, or create a contribution relation.
Show 5 (support-sensitive Method eligibility).
Keep the exact admitted Methods, row editions, and grouping basis from §0.5. In a gated variant, the matching G.4 task map makes AC_InputConditionGate-E1 from G.4 §5 applicable to ThresholdTrendReviewMethod-E2. Consume that clause’s value and threshold rather than define either in G.5. For the two-condition case there, the returned fail excludes that row from the assurance-gated set. Replacing its joint probability with minimum would wrongly retain it. An otherwise admissible row stays in the set under its own declared eligibility basis; if none survives, return the existing abstain or escalation outcome.
If the dependence model is missing, use the clause’s unknown branch rather than pass by a high F. In the ordinary, non-assurance question of §0.5, both grounded rows still form the unordered Shortlist. A formal proof, a limited complementary study, and an overlapping contrary result can remain separate support with their limitations; neither weak additional evidence nor the absence of an unjustified common score automatically removes a Method. A defeated necessary premise still changes the eligibility that actually relies on it.
Show 6 (one calibrated correspondence, two receiving uses).
Use G.7 §4.5’s VehicleTransportOrder row from C.3.3 to select among independently admitted Methods for a transport review. The task’s applicability comparison uses only the preserved transport/passenger order and explicitly ignores propulsion. When receiving admissibility, fresh target classification of the subject vehicles, the bounded comparison conditions and matching reliance pass, that row can support Method eligibility under the task’s other rules. Keep those premises with the result.
For a second task that requires battery-health evidence, the same CL^k=2 row loses the necessary EV distinction and cannot support that criterion. Return the missing battery premise through the existing unknown/abstain or evidence-request branch. Raising a scalar summary or citing a waiver does not restore that distinction. G.7 reports the correspondence; the exact TaskSignature and receiving rules decide each selection use.
G.5:6 - Bias-Annotation
These biases can affect dispatch:
-
Monoculture bias (single Tradition dominance by default). Mitigation: rows retain an explicit eligibility basis and conditional assurance expectations; public registration adds its stronger records; selection follows its declared outcome contract under partial orders; identify method-specific policies by their pins instead of hard-coding them as defaults.
-
Hidden scalarisation bias. G.Core preserves the comparator result and the commissioned selector outcome. Identify the governing dominance regime and any applicable ranking policy.
-
“Tool equals method” bias. Mitigation: notation independence and prohibition of tool keywords in core registry and eligibility fields; tool choices are outside the core.
-
Cross-sense leakage bias. Mitigation: when expressions have distinct source-local meanings, require exact F.17 endpoint senses, an obtaining F.9 Bridge, a separate C.2.1 bounded-use proposition, and the matching A.10 or B.3 reliance branch; keep loss and CL visible where material. Crossing pins and bundles remain audit or publication references and cannot make an implicit crossing admissible.
-
Survivorship bias in refresh. Mitigation: RSCR triggers are typed and id-based; freshness, decay, and telemetry deltas are first‑class causes with canonical ids.
G.5:7 - Conformance Checklist (normative)
| ConformanceId | Statement |
|---|---|
CC‑G5‑CoreRef | Core conformance bridge. G.5 is conformant only if the effective G.Core obligations referenced by G.5:4.1 (GCoreLinkageManifest) are satisfied (after profile and set expansion plus explicit deltas). |
CC‑G5.0 | Core standards SHALL remain notation‑independent; vendor or tool keywords are forbidden in registry, eligibility, assurance, or selector‑kernel obligations (E.5.*). |
CC‑G5.1 | Every reusable MethodFamily row SHALL fix its exact members, grouping basis, eligibility/comparison basis and selection-changing pins in one immutable edition. Intentional public registration additionally declares EligibilityStandardRef using CHR/CAL terms and applicable edition pins. Both branches remain notation-independent. |
CC-G5.2 | Selection SHALL be a pure function of TaskSignatureRef, exact method- and generator-family row refs, any conditionally current exact TaskMapRef, and pinned policy or edition refs; side effects are limited to emitting DRR and SCR pins, telemetry triggers, and RSCR triggers (no hidden mutation of constraint-bearing spec refs). |
CC‑G5.3 | Delegated (ID‑continuity) plus F.9 use boundary. When a selector use relates expressions with distinct F.17 source-local meanings, it MUST resolve the exact cells, an obtaining F.9 Bridge, a separate C.2.1 <u,d,r,t,polarity> proposition, and the matching A.10 or B.3 reliance branch. G.Core crossing visibility and penalty-assignment semantics still apply. Delegation targets: CC‑GCORE‑CROSS‑1, CC‑GCORE‑PEN‑1. Pins alone MUST NOT establish the Bridge, use, reliance, or actual selector application. |
CC‑G5.4 | Governing rule for DefaultId.GammaFoldForR_eff. An R or R_eff composition MUST use a justified receiving quantity, input meanings and scales, dependency model and operation under B.3/C.2.2; cite contributors and pin the actual model/policy. Neither a universal minimum/maximum or best-source cap, an ungrounded F-to-R conversion, nor monotonicity and boundedness alone supplies that model. Without a justified common model, retain separate support and a bounded synthesis. If an actual G.4 gate requires the missing quantity, follow that clause’s unknown behavior; a calculated threshold failure remains a failure. G.4 owns acceptance conditions and thresholds. Ordinary selection with no assurance condition remains usable without a new R calculation. |
CC-G5.5 | Ordinal scales MUST NOT be averaged or subtracted; any aggregation or comparison must respect CHR scale typing and admissibility constraints, including CSLC where applicable. |
CC‑G5.6 | Public registration. Intentional registration of a MethodFamily or GeneratorFamily under a stable public registry identity SHALL publish that identity to UTS with the applicable naming and lexical-continuity discipline. Project-local reusable rows keep the immutable member/basis contract of S1/S1′ without this added duty. Visibility alone does not select the public identity contract; actual audience availability is a separate E.24.PUB occurrence. |
CC‑G5.7 | Conditional. If G.5:Ext.EELog is present, exploration MUST be budgeted under the pinned exploration and exploitation log policy; probe outcomes MUST feed refresh through canonical RSCR trigger kinds. |
CC‑G5.8 | Actual CG-Frame gate enforced. When selection uses a CG-Frame gate, retain its pinned CG-Spec.MinimalEvidence requirements for the cited characteristics. Failed requirements reject or abstain under that gate’s rule; missing required input remains unknown/abstain and cannot pass. Registering a local grouping without this selection/gate use does not activate that gate. |
CC-G5.9 | Delegated (ID-continuity). Set-return semantics are pinned through G.Core. Delegation target: CC-GCORE-SET-1. Candidate ordering MUST be admissible over typed traits and admissibility constraints. If only a partial order is available, selection MUST return one declared selector outcome, for example one SetResultOutcome with Shortlist or RankedShortlist, one HandoffOutcome with SpecialistHandoff, or another pinned outcome result, with no forced totalisation via inadmissible scalarisation. |
CC-G5.10 | SCR completeness. SCR MUST enumerate Gamma-fold contributors when used, referenced constraint-bearing spec editions, the source and evidence citations used in gating and rationale (PathId or PathSliceId when graph citations are used or independently required), and MinimalEvidence gating verdicts by lane and carrier when such gating is relied upon. |
CC‑G5.11 | Delegated (ID‑continuity). Tri‑state eligibility and acceptance semantics plus unknown handling are pinned through G.Core. Delegation target: CC‑GCORE‑GUARD‑1. (Includes the rule that degrade(...) is expressed through a pinned FailureBehavior or SoS‑LOG branch id, not as a fourth status.) |
CC-G5.12 | Applicability of a selected method set. A selected method-set result MUST state the exact task, exact row refs or other exact members, grouping and selection basis, truthful outcome kind, ordering, selection or inclusion conditions, and the named next use. Add ClaimScope, selected A.2.6 slices, a validity or evaluation window, source or scheme editions, intended-use restrictions, counterexamples, evidence pins, and transfer or change conditions only when they change eligibility, selection, applicability, or a receiver’s justified reliance. Omitting any such action-changing value fails this check and reopens the affected result; values that change no current action stay out. Coverage, descriptors, and distance may guide a named search for omissions but MUST NOT establish broader applicability. Apply E.24.UK or A.8 only to a separate claim about a durable kind or kernel placement; neither claim widens the selected set’s applicability. |
CC‑G5.13 | Conditional. If the selector consumes admissibility or maturity records (e.g., through G.5:Ext.SoSLOG), it MUST NOT recompute thresholds; it consumes pinned admissibility ledger rows and cites clause and rung ids in audit pins. |
CC‑G5.14 | Φ(CL) and Φ_plane discipline. If crossing or plane penalties are applied, the active penalty policy ids (e.g., Φ(CL), Φ_plane) MUST be explicit in audit pins, and the pinned policies MUST satisfy the monotone and bounded requirements asserted by their cited constraint-bearing spec refs and be published through those same cited spec refs (e.g., CG‑Spec). SCR MUST record the policy id in use; penalty assignment semantics remain pinned through G.Core. |
CC-G5.15 | Unit and scale admissibility MUST be established via CSLC (A.18) before any aggregation or Gamma-fold; unit and scale mismatches are a fail-fast defect. |
CC‑G5.16 | Hidden thresholds are forbidden. Thresholds live in explicitly pinned acceptance or eligibility policy records, not in selector prose, LOG shells, or code. |
CC‑G5.17 | ReferencePlane MUST be declared (pinned) for any claim that is used in dispatch, and the selector’s audit records must cite it (including plane‑crossing pins when applicable). |
CC-G5.18 | Numeric comparisons and aggregations used by dispatch MUST cite an admissible, edition-pinned comparator or spec publication (as provided by the constraint-bearing spec refs); inadmissible mixes of scale types are forbidden. |
CC-G5.19 | Conditional (QD). If G.5:Ext.NQD is present, the required QD telemetry triple (quality, diversity, and QD summary) MUST be computable and ready for emission under the pinned descriptor and distance definitions and archive policy, without redefining their semantics in G.5. If actual publication is current, use E.17 for a source-backed face and return to source and E.24.PUB for the publication occurrence and audience availability. |
CC‑G5.20 | Conditional (QD). QD and illumination summaries are treated as telemetry unless explicitly promoted by a pinned acceptance or policy record; the selector must record the promoting policy id in audit pins. |
CC-G5.21 | Conditional (Archive and QD). An archive use cites its retention/insertion policy and the editions needed to reproduce the retained set. Descriptor, distance and related method definitions are required when that policy uses them, including a selected QD or distance-based branch. |
CC‑G5.22 | Conditional (QD). Twin‑naming discipline for descriptor vs plain space (if used) must be respected (distinct objects; no aliasing). |
CC-G5.23 | Default rule for DefaultId.PortfolioMode. The selector MUST expose PortfolioMode with values Pareto or Archive, with default = Archive, and echo it in DRR and SCR records and declared selector results when not explicitly overridden by pinned policy or TaskSignature. The default is a retention and evidence-preservation policy, not a public selected-set label, not a dominance default, and not a substitute for SetResultFamily. Epsilon-fronts are allowed as local decision aids under CG-Spec when explicitly pinned. |
CC-G5.23a | Parity-run publication. If parity harness is in use, a selector or generator MUST publish a parity run and ParityCard to UTS (see G.9). This obligation remains mandatory irrespective of dominance policy or PortfolioMode policy. |
CC‑G5.24 | Conditional (Open‑Ended). If G.5:Ext.OpenEndedFamilyWiring is present, the selector MUST return declared sets of {Environment, MethodFamily} pairs as set‑valued outcomes under explicit pins. |
CC‑G5.25 | Conditional (Open‑Ended). In Open‑Ended mode, TransferRulesRef.edition is mandatory and MUST be visible to telemetry and RSCR triggers. |
CC-G5.26 | Conditional (Archive and QD). Within any archive niche or cell, ordering and tie-breaks MUST remain admissible over compatible scales; inadmissible mixed-scale weighted sums are forbidden. |
CC‑G5.27 | If the selector cites any GateCrossing, the corresponding CrossingBundle publication MUST be present and conformant; missing or non‑conformant CrossingBundle blocks downstream consumption. The bundle packages already governed crossing evidence for that named use; it MUST NOT create the F.17 endpoints, F.9 Bridge, bounded-use proposition, A.10/B.3 reliance, gate decision, authorization, or actual selector use. |
CC‑G5.28 | Default rule for DefaultId.DominanceRegime. DominanceRegime SHALL default to ParetoOnly. Any inclusion of additional telemetry dimensions into dominance (e.g., illumination) requires an explicitly pinned acceptance or policy record and must be recorded in audit pins. Parity‑run publication (CC‑G5.23a) remains mandatory irrespective of dominance policy. |
CC-G5.29 | Conditional (QD and Open-Ended). Any telemetry event that materially changes an archive state or retained-set state MUST identify the exact changed state and affected-use scope, log PathSliceId when graph-scoped or independently required by the receiving contract, and retain the active policy id and active editions of the relevant definition pins (DescriptorMapRef.edition, DistanceDefRef.edition, and TransferRulesRef.edition when applicable) and expose them to RSCR triggers. |
CC‑G5.30 | No Strategy minting. Within G.5, “strategy” is a policy‑bound composition template; the pattern SHALL NOT mint a durable U-kind named Strategy (E.10 and E.24.UK discipline). If a stable reference is needed, publish composition and policy ids (e.g., UTS entries) rather than minting a universal kind. |
CC‑G5.31 | Blocked or empty selection. Retain the failed or unknown conditions and their basis. Abstention has no actual Select application or SelectionSlot binding; an actual Select may return an empty set only under its declared criteria. An ActionHint names a justified continuation when one is available and needed. When no further action is warranted, return the stop with its scope and grounds. Carry the editions and policy pins that the selected continuation uses. |
CC‑G5.32 | Parity‑run publication and admissible roll-ups. If parity harness is in use, parity publication is required per CC‑G5.23a (ID‑continuity). Any scalar roll-up or summary view MUST be admissible under CG‑Spec (no mixed‑scale sums), and published views must preserve set‑return semantics (no single‑score leaderboards as authoritative outputs without an explicit, admissible comparator publication). |
CC‑G5.33 | Conditional (bounded specialization). When the selection question is acquisition of usable specialization on a declared TaskFamilyRef or TaskSignature, selector outputs SHALL emit TaskFamilySpecializationProfile@Context or cite equivalent evidence under C.22.1. In either form, identify the setup or participant whose performance was checked, the assistance available during the check and in the intended work, the threshold and prior exposure. Give the observed time and budget to threshold or the bounds the checks support. Include the grounds for any claimed post-threshold efficiency, transfer, retention, downside or specialization entry. |
CC‑G5.34 | Selected-set result kind. When SelectorOutcomeKind = SetResultOutcome, the public result kind MUST be explicit. Use Shortlist for unordered alternatives retained for later choice, RankedShortlist only when the result orders those alternatives, and JointUseSet only when every named member is included for one named use. ChoiceSet MUST NOT silently replace the public result kind. |
CC‑G5.34a | Selector outcome typing. Declared selector results MUST state SelectorOutcomeKind. SetResultFamily is required only when SelectorOutcomeKind = SetResultOutcome; HandoffKind is required only when SelectorOutcomeKind = HandoffOutcome. Non-set outcomes MUST NOT masquerade as one public selected-set label. |
CC‑G5.35 | Result-content closure. Any declared selector result MUST state the SelectorOutcomeKind, applicable public result kind, retained members or keyed joint-use entries, ordering, named use and inclusion conditions when required, and basis pins directly in the emitted result rather than relying on upstream C.11, C.19, or C.24 notes. |
CC‑G5.36 | Neighboring-pattern boundary. If the current question is still local choice among already-available options, pool policy over still-live candidate lines, or enactment planning after choice, a G.5 use MUST consume the result produced by applying C.11, C.19, or C.24 rather than restating those patterns as if declaring selector-facing content decided the upstream matter. |
CC‑G5.37 | Derived tradition-view result discipline. If the selector emits one result through a derived tradition view such as TraditionFront or TraditionArchive, it MUST keep the declared base SourceSetFamily explicit, keep SoTAPaletteDescription recoverable through BasePaletteRef, and MUST NOT let the derived view become the default meaning of Tradition, TraditionPalette, or the base palette. |
CC‑G5.38 | Causal method dispatch declarations. If method selection involves causal methods, each compared method MUST declare causalMethodUseClassification as observational predictor, intervention optimizer, counterfactual strategy, causal fairness estimator, causal-RL policy, or simulation-only method, and MUST carry causalUseSupportResultRef and the cited result’s verdict when it consumes C.28 causal-use support rather than treating method dispatch as causal certification. |
CC-G5.39 | Registry grounding, edition, and grouping boundary. Every consumed method-family row MUST be addressed by one exact MethodFamilyRowRef whose immutable edition resolves its non-empty MethodRef[] to exact A.3.1 Methods; every consumed generator-family row MUST use one exact GeneratorFamilyRowRef and resolve non-empty exact generator refs under their subject patterns. Each row edition cites the independently established classification, membership relation, or explicit project-local grouping criterion used by this selector. A row, id, label, description, family card, eligibility or maturity record, policy, evidence pin, shortlist, or publication MUST NOT create a member or membership fact. Missing grounding or an unresolved row edition blocks that row’s family use. |
CC-G5.40 | Composition and selected-structure boundary. A composition shape MUST remain a template unless one already identified A.3.1 Method separately passes B.1.5’s complete composite-method qualification. An organization that does not constitute one Method MUST be consumed as an A.22 U.Structure only after all four A.22 identity discriminators are present. A template, A.3.2 description, registry row, selector outcome, diagram, label, or notation MUST NOT create either governed object or its underlying relations. |
CC-G5.41 | Declaration, actuality, performer, result and publication boundary. A registry, selector, policy, template, shortlist, DRR, SCR, telemetry or publication-content declaration MUST NOT be treated as an A.13 performer core, dated Work, F.6 attribution, actual A.6.1 Select application or binding, domain-result truth, C.2.1 result episteme, A.10 evidence-provenance path, B.3 assurance claim, authorization, or E.24.PUB publication occurrence. Every claimed precise performer MUST first have the A.13 core; when dated Work is asserted or required, A.15.1 MUST admit it independently; F.6 enters only for a current exact assignment-bound attribution through the same obtaining assignment. Every other claimed actual object or relation MUST be recovered under its subject pattern; missing actuality blocks only that stronger claim. |
CC-G5.42 | Crossing completeness. A selector use that relates expressions with distinct source-local meanings MUST NOT proceed from Bridge, CL, loss, registry, policy, CrossingAllowance, GateCrossing, CrossingBundle, DRR, or SCR pins alone. It requires exact F.17 endpoint senses, an obtaining F.9 Bridge, a separate C.2.1 bounded-use proposition, and the matching A.10 reliance disposition or B.3 assurance branch; authorization and the actual selector application remain separate. |
CC-G5.43 | Ordinary-use proportionality. A bounded selector run over already grounded rows MUST be usable from the exact task, immutable row-edition refs, members of the input candidate rows resolved under the applicable method- or generator-family rules, grouping bases, declared eligibility or comparison basis, truthful outcome, result members, ordering status, basis refs and next use. It MUST NOT demand a fresh registry build, crossing branch, A.10 reliance claim, B.3 assurance claim, stable public identity, or E.24.PUB occurrence unless that stronger object or claim is current. An actual selector use and its upstream mathematical comparisons retain their actual application bindings; dated selection or comparison Work is independently admitted when that Work is asserted or consumed. Missing conditional apparatus blocks only the stronger claim. |
CC-G5.44 | Joint-use membership integrity. A declared selector-result record with SetResultFamily = JointUseSet MUST name one bounded use, set ordering = unordered, and use keyed memberEntries with each exact memberRef present at most once. Entry order has no semantic effect; the result MUST NOT add an undefined per-member contribution or basis field; and any top-level members MUST be only the unique set projection of the entry keys, never an independently maintained list. Exact supporting content or claims remain in their own records and may be cited only among sufficient top-level basisPins. |
CC-G5.45 | Joint-use ontic and actuality boundary. A declared selector-result record with SetResultFamily = JointUseSet is well formed only if every exact framework-edition or other non-Method memberRef resolves under its existing identity and the record adds no MethodRef value or registry row merely for membership. Candidate-pool and excluded-candidate records, direct member relations, local choice, actual selection Work, publication availability, and access remain separate. |
CC-G5.46 | Operation-path integrity. A JointUseSet over non-Method members MUST be emitted through G.5-6 DeclareSetResult from exact already identified memberRef values and a current inclusion basis; it MUST NOT use RegisterFamily or G.5-3 Select. DeclareSetResult MUST NOT be treated as the upstream choice, an actual Select application, dated Work, persisted result episteme, or E.24.PUB availability occurrence. |
CC-G5.47 | G.4 TaskMap receiving boundary. TaskMapRef is required only when this selector uses G.4 CAL gates. Its immutable map edition MUST cite the same C.22 TaskSignatureRef supplied to selection, resolve one exact CALCharterRef and every cited clause, operator, flow, and evidence profile at its exact edition, and travel in result-basis and refresh pins. A mismatch or unresolved ref blocks that gated use. The map MUST NOT construct the TaskSignature, copy thresholds, duplicate acceptance semantics, or become mandatory for an ordinary selector with no G.4 gate. |
G.5:8 - Common Anti-Patterns and How to Avoid Them
-
Anti‑pattern: “Selector as a shadow spec.” Symptom: local acceptance or admissibility rules appear in selector prose or code, diverging from CN, CG, and CAL. Avoid: govern constraint semantics through
CNSpecRefandCGSpecRefplus pinned CAL records; keep G.5 core as a boundary. -
Anti‑pattern: “Implicit crossings.” Symptom: reuse across distinct source-local meanings is claimed from a shared label, Bridge or CL pin, registry row, policy, DRR or SCR line,
GateCrossing, orCrossingBundlewithout the required relation, use, and reliance facts. Avoid: resolve the exact F.17 endpoint senses; establish the F.9 Bridge; state the separate C.2.1<u,d,r,t,polarity>claim; require the matching A.10 disposition or B.3 assurance branch; and keep authorization and actual selector use separate. Materialize or cite a bundle only when its named downstream use requires that durable package. -
Anti‑pattern: “Hidden scalarisation.” Symptom: partial orders are flattened into single winners “for convenience”. Avoid: return declared sets; make dominance regimes explicit; keep telemetry report‑only unless promoted by explicit policy.
-
Anti‑pattern: “Method specifics in the selector head.” Symptom: QD, OEE, or preference models become mandatory for basic dispatch. Avoid: keep them in
G.5:Ext.*blocks with explicit pins andUses. -
Anti‑pattern: “Churn by meaning.” Symptom: a continuing family id silently resolves different members, grouping basis, or selection pins after a row changes. Avoid: keep the lineage id only for the continuing declared grouping, publish a new immutable row edition, and carry its exact row ref through selection, result basis, refresh, and deprecation notices.
-
Anti‑pattern: “Result declaration hidden in upstream reasoning.” Symptom: the retained alternatives or all-member result exist only as one implication inside
C.11,C.19, orC.24, whileG.5never names the declared result kind. Avoid: declare the selected-set result directly, with its result kind, applicable members or keyed entries, ordering, named use where required, and basis pins instead of leaving it implicit upstream. -
Anti-pattern: “Shortlist used for complementary members.” Symptom: every named member is needed for one use, but the result calls them alternatives in a
Shortlist. Avoid: useJointUseSet, name the joint use, and key one entry per exact member; keep direct member relations and actual selection separate. -
Anti‑pattern: “Declared result missing required content.” Symptom: a result is named but omits its ordering status, basis or information required for its kind, such as the members of a shortlist, the keyed entries and use of a joint-use set, the content of a handoff, or the reason for abstention. Avoid: state the result kind, ordering status and basis pins, with the members, named joint use, handoff content or blocking condition required for that outcome.
G.5:9 - Consequences
- Auditable plurality. Multiple Traditions can co-exist without forced semantic flattening; dispatch remains explainable and evidence-pinned.
- Core stability. Universal invariants are pinned through
G.Core; method innovation and generator innovation do not churn the selector head. - Evolvability. Registries allow growth, retirement, and refresh with typed RSCR causes and explicit payload pins.
- Composability. Strategy templates and fallbacks remain admissibility-checked and portable across implementations.
- Recoverable result content. A downstream reader can distinguish a shortlist or joint-use set from a handoff, abstention or escalation, and recover the grounds for the declared outcome.
G.5:10 - Rationale
- Why registries? Stable row editions preserve the grouping and eligibility basis for later dispatch. Assurance records enter when required by the declared use or public-registration contract.
- Why separation via Extensions? QD, OEE and preference-learning have different method semantics. Extensions let a selector use those semantics only for the task that needs them.
- Why name the result? A comparator can establish that two candidates are incomparable under the stated criteria. Returning both as a shortlist needs the candidate domain and the selector’s rule. A handoff, abstention or escalation has its own grounds.
- Why explicit defaults with one declared source? Indexing each default by its governing definition lets consumers use and revise the same rule.
- Why selected-set result declaration here? Once the current question is to state retained alternatives or an all-member result for downstream use, the selector should declare that result directly instead of leaving it implicit in local choice, pool-policy, or enactment notes written for other purposes.
- Why
JointUseSet? A shortlist preserves alternatives for later choice; an all-member result says that removing one member changes the result for the named use.JointUseSetis a localSetResultFamilyvalue using the existing member identities.G.5-3 Selectmay emit it only over exact Method candidates admitted through that kernel.G.5-6 DeclareSetResultcovers already identified members with a current inclusion basis, including framework editions, without retyping them as Methods or asserting a relation among them.
G.5:11 - SoTA-Echoing — return what the receiving use can actually select
Practice question. What should a dispatcher hand over when several candidates remain useful but the declared comparison does not justify one winner? G.5 returns the admissible alternatives with their ordering status and basis. A single-winner procedure chooses one best candidate under a declared objective and returns it ready for use. That procedure is efficient when the objective settles the choice; it loses a needed alternative when an undeclared scalar objective is substituted for an unresolved trade-off.
The scikit-learn GridSearchCV documentation, refit and best_estimator_, supplies the concrete default. It can choose one estimator by a scorer or a custom rule over the evaluation results; multi-metric evaluation still needs a specified refit choice. Adopt that explicit-choice requirement when G.5 emits an ordered result. Reject reading that best-estimator interface as a warrant to invent the missing comparator. The library supports custom choices and exposes other candidate results; it does not require G.5 to discard them.
PS-AAS, Kostovska et al. (2023) supplies a substantive alternative-selection line: form a complementary portfolio for later algorithm selection, balancing coverage with the cost of a larger choice problem. Its experiments concern specified CMA-ES variants and BBOB tasks, not arbitrary Methods. Adapt the distinction between forming an eligible pool and selecting a member to Shortlist and RankedShortlist in §4.4b, S3 and the pump case in §0.5. This use of PS-AAS concerns result declaration. Its experiments do not establish a best portfolio algorithm for every task.
A different receiving use combines members. Split-Ensemble, Chen et al. (2024) supplies a concrete counterexample to treating every retained set as alternatives: its submodels serve complementary subtasks in one ensemble. Adapt that use distinction to JointUseSet and G.5-6 DeclareSetResult. G.5 states all-member inclusion; the paper’s trained ensemble does not establish compatibility or effective combined use for an arbitrary set of framework editions. Show 4 therefore keeps those claims with their own patterns.
In §0.5 both grounded pump Methods meet the task constraints, but no admitted comparator orders them. With the same candidate evidence, returning the pair as an unordered Shortlist preserves the receiving choice at the cost of one further decision. Naming the spectral Method “best” would require an additional criterion and its supporting comparison. If that criterion is supplied and actually orders the candidates, a ranked result is available. If the receiving use includes every exact member, declare that different membership meaning; a ranking does not establish it. These distinctions govern the outcome declarations and the positive, ranked, no-survivor and joint-use cases in §5.
Reopen when new evidence changes eligibility, a comparator supplies or defeats an ordering, the downstream use changes from alternatives to joint inclusion, or keeping the retained set costs more than the receiving decision can bear. A bounded handoff or abstain outcome remains available. None of these sources establishes actual selection Work, a Method’s identity, public availability or assurance merely from the emitted result.
G.5:12 - Relations
Builds on (normative): G.Core (core invariants + linkage discipline).
Uses (conceptual dependencies; cited via pins and ids):
-
Specification refs required by this result:
A.19.CN (CN‑Spec),G.0 (CG‑Spec). UseA.2.6only when aU.ClaimScopeor selectedU.ContextSlicechanges selection, applicability, or a receiver’s justified reliance; validity and evaluation windows and intended-use restrictions follow the same conditional boundary. -
Method identity and family grouping:
A.3.1for every exact selectableU.Method;A.3.2only for the same C.2.1 episteme that substantively describes one already admitted Method; and C.2.1 or the defining declaration or pattern for the family relation cited by a registry row. G.5 creates none of those source facts. -
Method composition and selected organization:
B.1.5for the complete composite-Method qualification,A.22for an independently selected organization that does not constitute one Method, andC.29when the current use needs an unresolved mathematical representation or transfer contribution under §4.2 and the mathematical-lens route below. G.5 consumes exact resulting references and does not construct them. -
Upstream object sets:
G.1 (CG‑Frame Card),G.2 (SoTA Pack),G.3 (CHR Pack), andG.4 (CAL Pack). C.22 alone constitutes the TaskSignature. When G.4 CAL gates are current, G.5 additionally consumes the exactTaskMapRefthat relates that sameTaskSignatureRefto one exact charter and the cited CAL declarations; otherwise the map is absent. -
Evidence and crossings:
G.6for EvidenceGraph citations;F.17for exact local senses;F.9for the direct Bridge; C.2.1 for the separate bounded-use proposition; andA.10orB.3for reliance or assurance. Add aCrossingBundleunderE.18or a GateCheck underA.21only when that named downstream use requires one. A G.7 calibration artifact remains a cited policy or evidence input; it does not define the Bridge, bounded use, reliance, or selector actuality. -
Planning and enactment boundary:
A.15.2identifies theU.WorkPlanused asplannedBaselineRef; A.15.3 defines any planned-filling rows kept inside that WorkPlan. G.5 does not redefine them. -
Actual selector use and result availability:
A.19.SelectorMechanismand A.6.1 for the actualSelectapplication and bindings; when dated selection Work is asserted or required, A.13 for each precise performer’s local-kind criterion, classification, same obtaining assignment, scope, situation, window, and evidence; A.15.1 for that independent Work admission; A.2.1 for the assignment species and occurrence; and F.6 only for a current exact assignment-bound attribution. When asserted by the account or consumed by its receiving use, A.2.4 governs evidence use, A.10 governs reliance and provenance, G.11 governs currentness, C.2.1 governs a persisted result episteme, B.3 governs assurance, the direct authority pattern governs authorization, and E.24.PUB governs publication. Upstream CPM applications retain their local bindings without requiring separately admitted comparison Work unless that Work is asserted. A root-family assignment reference, temporal overlap, or omission from short wording supplies no attribution and removes no world-side fact. G.5 declarations and records create none of those neighboring facts. -
Joint-use members outside Method dispatch: the direct identity pattern identifies every
memberRef;C.11supplies a local choice result when one is current; another accepted decision or governed inclusion basis may establish all-member inclusion; E.4.PFR states framework-edition dependency or pairwise compatibility separately;G.11supplies currentness; and E.17/E.24.PUB plus the applicable access-carrier pattern supply exposure and source return.G.5-6 DeclareSetResultconsumes the exact members and sufficient basis pins and emits only the selector-facing membership result. -
Causal-use method dispatch:
C.28when method selection involves causal effect, counterfactual comparison, causal fairness, causal policy, causal RL, or simulation-only causal-use claims. -
Optional Method or generator extensions through
G.5:Ext.*:C.18,C.19,C.23, plus extension-bearing patterns whose exact Part G admission relation is established when they add extra selector pins. -
Mathematical-lens use: apply
C.29when a selector input depends on a mathematical object or mapping whose use is not yet recoverable—for example, a comparator, distance, descriptor geometry, embedding, normalization, surrogate model, learned representation, QD archive descriptor, model-family label, or model-selection basis. For claim-bearing lens use, recover that object’s mapping mode, preserved or lost structure, and stop condition. ALensCandidateNotemay instead retain the recognition and next-action account whileCandidateMathObject?remains unresolved. The recorded result is a C.29 lens-use result; non-exhaustive examples include no lens use, a lens-candidate note, a one-line note, a mini-card, a full card, or a note naming the applicable pattern for the stated selector use. That result does not declare a selector result or its supporting records, such as the selected set, selector policy, registry row, shortlist, ranked shortlist, or selector evidence pins; useG.5for those objects and cite the exact references used.
Provides to: downstream uses such as G.6 audit citations, RSCR emission records with typed triggers and payload pins, and packs shipped through G.10. When a named use needs stable public identity, publish the required family ids, selector policy records, or selected-set identities—such as ShortlistId—to UTS under the applicable identity rule.
Coordinates with: C.11 for local choice results; E.4.PFR for direct framework-edition dependency and pairwise compatibility claims; G.11 for edition currentness; E.17 for a source-backed publication face and return to source; E.24.PUB for a publication occurrence and audience availability; C.19 for pool-policy records; C.32.P2S when a selected-set result declaration feeds architecture problem-to-structure carry-through; C.35 when a generated or discovered structure-bearing output is not yet a selector-facing result; C.24 for enactment-facing next-action records; and C.18 when a Front or Q-front is the source set for a G.5 use.
A Q-front stays a C.18 source set; it is not the emitted G.5 outcome or a SetResultFamily. The G.5 result states one admitted SelectorOutcomeKind; a set result also states Shortlist, RankedShortlist, or JointUseSet, and any public selected-set label resolves to that family.
Architecture discovery boundary: when a generated or discovered structure-bearing output is only a representation or carrier—for example, a description, query result, graph, cluster, or search trace—use C.35 before G.5. Use G.5 only when the live claim is declaration of selected-set result content with selector-policy and selected-set identity; stable public identity and actual publication remain conditional neighboring branches.
G.5:End
G.6 - Evidence Graph and Provenance Ledger: Citable Evidence-Provenance Paths
Type: Evidence and provenance pattern Status: Stable Normativity: Normative where conformance rows say so; examples and SoTA rows are informative guidance.
G.6:1 - Problem Frame
Use this pattern when a later user must cite, replay, audit or refresh a chain of established facts, such as a measurement, an acceptance judgement and a decision. Give that chain a stable address while preserving the evidence and rules for its members.
Primary EntityOfConcern. G.6 concerns an addressable provenance representation: one EvidenceGraph, its PathId or PathSliceId, and any ledger entry that makes the path replayable. G.6 governs path identity, slicing, citation and refresh of the representation. The represented facts remain governed by their subject patterns.
First useful move. Name the relied-on claim or bounded use, then list the exact object refs and direct relation refs needed to replay it. For each relation, identify the rule that governs it and cite the claim that it holds. Only then draw the path. Keep an unresolved relation as a gap; do not turn it into a graph edge asserted as obtaining.
What goes wrong if missed. Reading a graph as proof of the facts it depicts can make an unperformed method look like Work, a co-listed System or entity look like a participant or Work performer, a carrier look like a produced result, a measurement or verdict look like generic evidence, or a provenance edge look like the world-side relation itself.
What this buys. A downstream user can cite one path and recover its supporting facts, their qualifications and the direct relations that connect them. A changed premise can be traced to every dependent path within the established coverage.
When to use a different pattern. If the required fact or relation has not been established, use its subject pattern in §4.1 first. If only one local source-to-use statement is needed, stay in A.10.
Here path means a path in a descriptive provenance graph, not an action route.
G.6:2 - Problem
Large projects often need to cite a chain that crosses measurement, evaluation, aggregation, production, publication, and later use. Relying on the chain can lead to a wrong conclusion if a reader takes the graph as grounds for a fact that its subject rule has not established.
The common failures are:
- Edge-to-fact inversion. A drawn edge is treated as proof that work, participation, production, measurement, evaluation, or use occurred.
- Generic relation fallback. Labels such as
verifiedBy,validatedBy,measuredBy,producedByWork, orevidencesreplace the exact direct relation and its governor. - Result collapse. Subject result, result episteme, carrier, outcome, assurance, and later decision become one generic result node.
- Declaration-to-runtime collapse. A
MethodDescription, operation signature, policy, clause, or plan is read as an actual run and its bindings. - Hidden crossing. A path silently crosses context, reference plane, edition, source order, or currentness window.
- Incomplete refresh reach. A changed premise prompts a global rerun or a repair of only the first located slice because its actual dependent paths have not been identified.
G.6:3 - Forces
| Force | Tension this pattern resolves |
|---|---|
| Compact citation versus the represented facts’ governing rules | One path is easy to cite, but each represented fact and relation must remain with its exact governor. |
| Graph readability versus ontic force | Nodes and edges make a chain legible; their presence cannot make any represented relation obtain. |
| Result continuity versus result collapse | A path may connect measurement, evaluation, aggregation, and decision while preserving every local result and result episteme. |
| Reusable declaration versus performed occurrence | Methods, descriptions, policies, and clauses may be cited, but dated work and actual bindings remain separate. |
| Cross-context reuse versus hidden loss | Bridges, editions, time windows, source order, and currentness remain visible at the path slice that depends on them. |
| Refresh locality versus stale reliance | One changed premise may affect several paths; complete dependency coverage excludes unrelated paths without losing a dependent use. |
G.6:4 - Solution — cite independently governed objects and relations
Create an EvidenceGraph only after the relied-on claim or bounded use and its supporting objects have been recovered. In this declarative, addressable representation, each node record cites one independently governed object. Each asserted edge record cites one independently established direct relation. PathId, PathSliceId, and the provenance ledger add citation and refresh locality, not world-side facts.
G.6:4.1 - Subject-pattern map
| Represented claim or object | Subject pattern before G.6 represents it |
|---|---|
| Reusable method, generic participants, parameters, effects, and conditions | A.3.1 for the exact U.Method; A.3.2 for its U.MethodDescription |
| Independently admitted dated Work and refs to its performers; optional obtaining F.6 relation and assignment occurrence refs when the path expressly relies on that attribution; enactment, resources, and direct participation or binding facts | A.13 to identify each performer of the Work and A.15.1 for independent Work admission; F.6 and A.2.1 only when the path represents the claim that the Work was performed under the stated assignment; the exact direct participation or resource relation; and A.6.1 for operation-application bindings |
| Production or inception of an entity or episteme | one exact local A.15.PROD claim when its entry condition is met, or a direct subject predicate under its own pattern |
| Measurement result and its measurement-specific basis | C.16 |
| Acceptance-clause application or other runtime evaluation result | G.4 or the exact formal, conformance, diagnostic, causal, comparison, selection, gate, or decision governor |
| Work-resource aggregation result | B.1.6 |
| Durable episteme that states a local result | C.2.1; it remains distinct from the domain result |
| Outcome, later action, acceptance, gate passage, permission, or decision | its exact work and domain governor, including C.11 or A.21 when applicable |
| Source publication, carrier, copy, extraction, or publication occurrence | E.17 for a source-backed face and source return; E.24.PUB for an obtaining publication occurrence; the direct rule for each copy, extraction or source relation |
| Representation correspondence | C.29 |
| Bridge, congruence, loss, or cross-context transfer | F.9 |
| Transformation-flow structure distinct from performed work | E.18 and E.18.2 |
| First evidence/status use, provenance and bounded reliance, currentness, or assurance | A.2.4, A.10, G.11, or B.3 respectively |
If the subject pattern or relation cannot be recovered, record an unresolved gap in the path and leave the edge unasserted.
Do not add a local U.EvidenceRole or turn proof, measurement, benchmark, source, or status labels into system-role kinds. For any claim that a producer, verifier, laboratory, issuer, or maintainer participates, recover the exact direct relation, the participants it declares, and the place each actual participant fills. Other nearby facts—for example, a local system-role kind, assignment, Work occurrence, responsibility, authority, or permission—are separate and may be cited only when they independently obtain; none establishes participation. Do not infer that a passive laboratory or produced entity performs Work merely because the path cites it.
Work recovery and compact citation. Before G.6 represents dated U.Work, its subject account must already identify each performer of that Work under A.13 and establish the independent admission of that Work under A.15.1. Include an assignment occurrence and obtaining F.6 relation only when the graph path or receiving use expressly relies on the claim that the Work was performed under a stated assignment; any present attribution must use the same obtaining A.13 assignment. If a required Work or performer ref is absent, record a Work gap. If a required F.6 ref is absent or unresolved, retain the Work node and record an attribution gap rather than suppressing the Work. G.6 neither re-admits the Work nor retests assignment species, occurrence identity, holder, classification, predicate duration, or interval coverage. Merely listing an assignment beside Work establishes no relation between them.
G.6:4.2 - EvidenceGraph as a representation
An EvidenceGraph is a typed directed graph used for provenance citation and replay. It may project a dependency-closed slice of independently governed objects and relations.
Minimal graph fields:
EvidenceGraph:
EvidenceGraphId
ReliedOnClaimOrBoundedUseRef
BoundedContext
ReferencePlane
RepresentedNodeRecords
RepresentedRelationEdgeRecords
TimeWindowOrPolicy
SourceCurrentnessRefs
BridgeOrLossRefs
EditionOrPolicyRefs
GraphPathAddressingRule
C29RepresentationRefs
A node record is a projection:
RepresentedNodeRecord:
GraphNodeId
RepresentedObjectRef
ObjectKindAsGoverned
SubjectPatternLocator
ContextEditionOrTimeQualification?
RepresentationRef
Node records use the object kinds and subject patterns in §4.1. Every cited Work occurrence uses the independently established performer and A.15.1 Work refs specified there; add F.6 refs only when the path relies on the claim that the Work was performed under a stated assignment. Co-listing creates no relation among the cited objects.
An asserted edge is also a projection:
RepresentedRelationEdgeRecord:
GraphEdgeId
DirectRelationRef
DirectRelationKindRef
ActualParticipantRefs
SubjectPatternLocator
ObtainingClaimRef
ContextEditionOrTimeQualification?
RepresentationRef
Before the edge enters a relied-on path, the exact direct relation must already be established under its governor. The edge must reference each participant in the position it fills in that relation; adjacency, direction, shared identifiers, timestamps, source order, or visual layout cannot supply them. RepresentationRef points outward to the applicable C.29 correspondence when that correspondence is current.
G.6 defines no fallback core edge vocabulary. Legacy or display labels such as verifiedBy, validatedBy, measuredBy, producedByWork, derivedFrom, usesMethodDescription, citesSource or evidences are navigation prompts. Recover the exact direct relation and its governor before asserting an edge as obtaining.
G.6:4.3 - PathId and PathSliceId
A PathId identifies one claim-local path inside an EvidenceGraph. A PathSliceId identifies the same path under a declared time window, reference plane, bounded context, edition, bridge, policy, or selected object/relation subset.
Use this compact record:
PathCitationRecord:
ReliedOnClaimOrBoundedUseRef
EvidenceGraphRef
PathId
PathSliceId
BoundedContext
ReferencePlane
RepresentedObjectRefs
RepresentedDirectRelationRefs
SubjectPatternLocators
SourcePublicationAndCarrierRefs
C29RepresentationRefs
TimeWindowOrFreshnessPolicy
SourceCurrentnessRefs
BridgeOrLossRefs
EditionOrPolicyRefs
DownstreamWorkRef?
ExactDownstreamUseRelationRef?
A10RelianceDispositionRef?
NotCarried
UnresolvedRelationGaps
ReopenTrigger
NotCarried names the stronger claim or use at issue that the path does not establish. Recover the exact premise, reference, operation-argument, decision-use, or other direct relation for actual downstream use. If the use asserts dated U.Work, cite its independently admitted A.15.1 Work ref and A.13-qualified performer refs. Add attribution refs only when the use expressly relies on the claim that the Work was performed under a stated assignment. Path availability or citation alone does not establish that use.
G.6:4.4 - Provenance ledger
A ProvenanceLedger indexes PathCitationRecord entries for citation and replay. Its object and relation indexes let a user find paths that cite the same basis.
ProvenanceLedger:
LedgerId
EvidenceGraphRef
PathCitationRecords
RepresentedObjectIndex
RepresentedDirectRelationIndex
SourceOrderPolicy
CurrentnessPolicy
PrivacyOrDisclosureBoundary
RefreshScopeRule
Use a ledger when several downstream consumers need the same path family; a single local source-to-use account can remain in A.10.
G.6:4.5 - Refresh and source return
When a cited object, direct relation, governor, source, bridge, representation correspondence, edition, policy, time window, currentness result or reliance boundary changes, start at its node or relation-edge projection. Identify the exact changed claim or condition and the receiving use it may alter. Use the represented-object and direct-relation indexes to locate other PathId or PathSliceId values that consume the same basis. Treat the first slice as a starting point. Check for other uses of the changed basis before concluding how far the change reaches.
For each candidate path, recover the actual premise, evidence-use, operation-argument or other direct use under its governor. Follow that dependence while a receiving claim, action or required check can change. A shared calibration, causal assumption or source conclusion can affect several otherwise separate paths; include each affected use. Common storage, a citation or graph reachability alone establishes no such dependence. When a changed source requires discovery beyond an adequate existing index, use A.10.1 for its bounded search, direct-use test and coverage limits. An unsearched or unresolved branch is a limit on the conclusion, not an unaffected branch.
Retain independent results and their qualified paths. For an affected branch, return to the subject pattern to determine whether the available basis supports the same result, a narrower use, a changed result or a scoped gap. If a represented relation no longer obtains or its required support is unavailable, remove it from the relied-on path or mark that gap; retain the earlier account for its historical use where warranted. G.11 uses the resulting dependency scope to decide which refresh actions, if any, are needed. Updating a projection does not establish a replacement relation or perform that Work.
G.6:4.6 - Declarative representation discipline
EvidenceGraph, PathId, PathSliceId and ProvenanceLedger identify the governed accounts a reader can inspect. When a needed account is missing, return to its subject pattern.
| Current phrase or artifact | Required recovery before G.6 representation |
|---|---|
| method, protocol, algorithm, clause, or policy | exact reusable declaration; when the path cites dated Work, recover it separately under §4.1; when it cites actual operation bindings, recover them under A.6.1 |
| work trace, run, test, audit, measurement, or evaluation | independently admitted dated Work ref and refs to the performers identified under A.13 as described in §4.1; enacted Method, resources, exact direct participation facts, and A.6.1 binding facts remain separate; expose an assignment occurrence and obtaining F.6 relation only when the path expressly relies on the claim that the Work was performed under a stated assignment |
| produced carrier, model, report, or episteme | exact produced entity and either its subject-specific direct production relation, when the subject pattern declares one, or the one local A.15.PROD production-work or inception claim that the current use needs |
| reading, score, verdict, estimate, aggregate, diagnosis, or outcome | exact domain result and direct governor; distinct C.2.1 episteme when durably stated |
| publication, view, export, or graph rendering | exact source relation, E.17 source-backed face, E.24.PUB publication occurrence, and C.29 representation correspondence when each is current |
| evidence, provenance, currentness, reliance, or assurance | A.2.4/A.10, G.11, and B.3 under their separate entry conditions |
| later acceptance, gate, release, or decision | its local result and exact later-use relation under their direct rules; dated Work admitted under §4.1 when that occurrence is asserted |
G.6:4.7 - Extension wiring without core drift
Selector, benchmark, assurance, refresh or telemetry patterns may require additional pins in PathCitationRecord and cite PathId or PathSliceId. Each added record names its represented object or direct relation and governor.
G.5 may cite a path for selector explanation, G.9 for benchmark replication, G.11 for local refresh, and B.3 for an assurance input. Their selection, benchmark, currentness, and assurance results remain their own.
G.6:5 - Archetypal Grounding
G.6:5.1 - Measurement, acceptance, and decision
In this constructed case, a measurement team performs dated C.16 measurement Work with bindings for the pressure measurand, detector, calibration, model, input quantities and uncertainty propagation. A distinct C.2.1 episteme states the pressure measurement result.
An evaluation team later performs G.4 EvaluationWork, applying one declared acceptance clause through exact A.6.1 bindings and obtaining unknown; another C.2.1 episteme states that verdict.
A decision maker then performs C.11 decision Work, uses the verdict episteme through an exact premise relation and defers. For all three Work occurrences, the performer and Work references are independently established under §4.1.
G.6 may give this chain one PathId only after the measurement, work, binding, result, episteme, clause-application, premise, and decision relations are independently established. Its nodes keep raw detector output, indication, actual pressure, measurement result, verdict, and decision distinct. Its edges cite the exact relations; none produces the work, verdict, or decision.
G.6:5.2 - Resource aggregation
An engine programme has several C.16 resource measurements, dated test-run Work occurrences, exact phase and overlap relations, and a shared warm-up allocation rule. An aggregation team performs dated B.1.6 aggregation Work, applies ProgrammeResourcePolicy-v3 and obtains a typed resource vector with propagated uncertainty; a distinct C.2.1 episteme states it. The path uses the independently established performer and Work refs.
The G.6 path cites every measurement result and episteme, the work-set and overlap relations, the edition-pinned policy, aggregation work, aggregation result, sources, and representation refs. Epoch labels alone do not establish work-part relations. Warm-up energy allocation and uncertainty propagation are performed in the aggregation work; an emissions verdict remains a separate result.
G.6:5.3 - Produced model and benchmark use
A training team performs dated training Work with identified application bindings. When the account asserts inception or completion, it also cites one local A.15.PROD claim. An evaluation team separately performs benchmark-evaluation Work, applies its declared method and dataset edition and obtains a result under the benchmark’s direct governor. A C.2.1 episteme states that result. The path cites independently established performer refs for both Work occurrences. E.17 governs the model card’s source-backed face and return to the selected claims, E.24.PUB their actual audience availability, and C.29 any consumed representation correspondence. G.11 supplies currentness when later use depends on edition or freshness.
A G.6 PathSliceId may cite that dependency chain for replication. The graph does not infer training from the model’s presence, participation from a roster, evaluation from the protocol, superiority from the score, or deployment permission from the model card.
G.6:5.4 - Dashboard status cue
A dashboard cell shows Ready. F.10 governs the status-use classification; A.10 recovers the source and provenance for bounded reliance, with query Work, currentness, and a live rival explanation when those facts affect the claim. G.6 is entered only when a downstream audit or release package needs a stable path through those already established relations. The visible cue, graph path, and ledger row establish neither gate passage nor release.
G.6:5.5 - One calibration basis, several receiving paths
In a constructed inspection account, Path-P1 connects measurement result M1 to acceptance result V1, and Path-P2 connects M2 to V2. Both measurements use the same calibration qualification K through independently established relations. A later selection S uses both acceptance results. Path-P3, stored in the same publication, uses a different calibration and supports an unrelated result.
A correction to K first appears in P1’s projection. The ledger’s indexes also locate P2, and the receiving-use relations lead from V1 and V2 to S. Reconsider both branches and S under their subject patterns. P3 stays outside this reach because it does not consume the changed qualification. P1 and P2 cite the same K. Counting both paths supplies no independent corroboration. If the index cannot cover another receiving store, state that discovery limit and use A.10.1 when the receiving question needs that search.
The resulting G.6 account identifies the affected paths and their exact dependencies. G.11:5 continues this case through different branch outcomes and selection of the needed work.
G.6:6 - Bias-Annotation
| Bias | Guard |
|---|---|
| Graph-authority bias | A node or edge represents an object or direct relation only after that object or obtaining relation has been independently established under its governing rule. |
| Generic-edge bias | Reject fallback verifiedBy, validatedBy, measuredBy, producedByWork, and evidences relations; recover the exact direct relation. |
| Result-node bias | Keep subject result, result episteme, carrier, outcome, assurance, and later action distinct. |
| Declaration-runtime bias | A method, description, policy, clause, signature, or plan establishes no occurrence or actual binding. |
| Provenance-as-truth bias | Origin and history support only their named bounded claim; provenance is not truth, safety, approval, or assurance. |
| Path-as-workflow bias | Graph path identity supports citation and refresh; actual work and transformation flow retain their subject patterns. |
| Ledger-process bias | The ledger contains replayable provenance records, not campaign status, review proof, or work-progress notes. |
G.6:7 - Conformance Checklist
| ID | Check | Repair if missing |
|---|---|---|
CC-G6-01 Exact use | Is one relied-on claim or bounded downstream use named? | Name it, or stay in local A.10 source recovery. |
CC-G6-02 Object projection | Does every node cite an exact independently governed object, kind, governor, qualification, and representation ref? | Recover the object or record an unresolved gap; do not mint a graph-only world object. |
CC-G6-03 Relation prerequisite | Does every asserted edge cite one exact direct relation, its actual participants, governor, obtaining claim, and context? | Establish the direct relation first or remove the edge from the relied-on path. |
CC-G6-04 No fallback edge | Are legacy or display labels prevented from acting as universal relations? | Replace each with the exact formal, measurement, work, production, publication, representation, provenance, temporal, status-use, or later-use relation. |
CC-G6-05 Work boundary | For each represented Work, does the path cite its independently admitted A.15.1 Work ref and its refs to the performers identified under A.13? Are assignment occurrence and F.6 refs included only when the path expressly relies on the claim that the Work was performed under a stated assignment, with a missing attribution recorded as a gap rather than loss of the Work node? Are Method, MethodDescription, resources, direct participation, and A.6.1 bindings separate? | Use A.13 and A.15.1 for the already-established performer and Work. Cite A.2.1/F.6 only when the path expressly relies on that attribution, A.6.1 for actual operation bindings, and the exact direct relation for every other participant claim. |
CC-G6-06 Result boundary | Are produced entity, subject result, result episteme, carrier, outcome, assurance, and later action distinct and independently identified under exact predicates? | Handle each through the exact predicates and assertions located in A.15.PROD, the domain result pattern, C.2.1, E.17/E.24.PUB/C.29, B.3, or the later-action source. |
CC-G6-07 Source and representation | Are source publication, carrier, copy/transform chain, and C.29 correspondence explicit when current? | Recover those relations before treating the graph rendering as source truth. |
CC-G6-08 Time and crossing | Are bounded context, plane, window, bridge/loss, edition, policy, source order, and G.11 currentness visible where they limit use? | Add the exact refs or narrow/block the path slice. |
CC-G6-09 Provenance and use | Are A.2.4/A.10 evidence/status use, A.10 provenance/reliance, downstream work, and exact use relation separate? | Recover the direct use; path citation or membership is not actual reliance. |
CC-G6-10 Ledger boundary | Does the ledger index only already established objects and relations, with NotCarried, gaps, and local reopen triggers? | Remove process status, generic result fields, and fact-creating language. |
CC-G6-11 Affected reach | Does the refresh scope include every established dependent use that can change within the declared coverage, including uses of a shared premise on other paths? | Follow §4.5; retain independent paths, expose unresolved coverage and return each affected judgement to its subject pattern. |
G.6:8 - Common Anti-Patterns and How to Avoid Them
| Anti-pattern | Why it fails | Repair |
|---|---|---|
| Edge as fact | Drawing or storing an edge is mistaken for an obtaining relation. | Establish the exact direct relation under its governor, then cite it through a representation record. |
| Universal evidence edge | verifiedBy, validatedBy, measuredBy, producedByWork, or evidences absorbs several relation families. | Replace the label with the exact formal, measurement, work, production, source, use, or other direct relation. |
| MethodDescription as run trace | Generic declarations acquire actual participants, time, or results by graph membership. | Cite one independently admitted dated Work ref and refs to its performers identified under A.13 through §4.1. Keep Method enactment, resources, direct participation, and A.6.1 bindings separate; expose an assignment occurrence and F.6 relation only when the path expressly relies on the claim that the Work was performed under a stated assignment. |
| Generic result node | Measurement, evaluation, aggregation, episteme, outcome, and decision collapse. | Keep each local result under its domain governor and each durable assertion under C.2.1. |
| Provenance as result or assurance | A path or ledger row is read as truth, currentness, safety, permission, or acceptance. | Use A.10, G.11, and B.3 under their entry conditions, and state the exact local result under its applicable predicate and pattern. |
| Citation as actual use | A downstream record cites a path and is assumed to have used it. | Establish the exact premise, reference, argument, or decision-use relation; recover dated Work separately under §4.1 when the claim asserts that occurrence. |
| Workflow overread | A declarative path becomes a method or action route. | Handle Work under A.15.1 and transformation-flow structure under E.18; limit G.6 to representation and citation. |
| Global refresh or first-slice stop | A changed premise either reopens every graph or leaves another dependent path unchanged. | Recover all actual dependent uses within the covered scope under §4.5; retain independent paths and expose gaps. |
G.6:9 - Consequences
Benefits:
- downstream records cite evidence-provenance paths without copying evidence tables;
- source, bridge, policy, edition and time changes reopen their full affected dependency scope while preserving independent results;
- evidence, assurance, causal use, status, gate, work, and publication claims stay in their subject patterns;
- scoped references let downstream users replay the provenance while limiting what they disclose.
Costs:
- path identity, node typing, and source-currentness refs add overhead;
- graph paths can look like routes unless declarative representation discipline is kept visible;
- users must resist treating one complete path as a complete downstream decision.
G.6:10 - Rationale
A.10 recovers one relied-on claim, its source/provenance account and bounded reliance. G.6 adds stable path identity, slicing and shared citation. Its indexes help locate uses of a changed common basis; A.10.1 supplies further source-change discovery when needed, and G.11 selects refresh from the resulting scope.
That representational gain does not justify a second ontology of evidence edges. Work, participants, products, subject results, result epistemes, outcomes, sources, provenance, currentness, and later uses already have direct governors. G.6 therefore projects their exact refs and direct relations, and C.29 governs the representation correspondence when current. This makes a complex chain readable without allowing graph topology to create facts.
The ledger is likewise an index over established provenance, not a result store or process log. Show missing evidence for a relation as a gap. Redrawing the path cannot supply that evidence.
G.6:11 - SoTA-Echoing
The table shows what G.6 adopts from the publishers’ editions current on 30 July 2026. Reopen an affected source-use decision when a new specification, Recommendation, maintenance change or replacement alters a relied-on identity, provenance, validation or use condition; follow the dependent-use return in :4.5.
| Exact source and source-use decision | G.6 use | Rejected overread | Direct source-change entry |
|---|---|---|---|
| W3C PROV-O, Recommendation 30 April 2013 — adapt qualified provenance descriptions and stable entity/activity/agent references only as a representation discipline for exact FPF objects and direct relations. | RepresentedNodeRecord, RepresentedRelationEdgeRecord, the measurement-to-decision case, and CC-G6-02/03 require every node and edge to cite an independently governed object or obtaining relation with its governor and qualification. | A PROV-shaped class, activity, agent, qualified association, or derivation does not establish FPF work, participation, production, result, truth, currentness, or later use. | Start with §4.2’s node/edge rules, the measurement-to-decision path, and CC-G6-02/03 if PROV-O’s qualified-relation contract changes. |
| C2PA Content Credentials Technical Specification 2.4, April 2026 — adapt asset/manifest identity, claim generator, assertions, ingredients/actions, signature validation, trust policy, and specification version for claim-bound content provenance. | PathCitationRecord carries source publication/carrier, C.29 representation, edition/policy, currentness, and NotCarried; the produced-model case and CC-G6-07/08 retain the exact content carrier, transform chain, trust regime, and version. | A valid manifest, visible Content Credential, ingredient chain, or authenticity mark does not establish truth of the represented world state, authorship beyond its exact assertion, work, safety, permission, or adequacy. | Start with those PathCitationRecord source/version fields, the produced-model carrier slice, and CC-G6-07/08 when C2PA changes manifest/assertion identity, validation, trust, or version semantics. |
SLSA specification v1.2 with in-toto Attestation Framework v1.2 and Statement/v1 — adapt artifact subject, predicate type, producing context, inputs, authenticated envelope, verifier expectation, and versioned attestation separation. | The produced-model/benchmark path names training work, produced model edition, dataset/method edition, benchmark work/result, source inputs, publication/carrier, verifier context, and currentness; CC-G6-07/08 keep those refs replayable without one generic attestation edge. | A signed statement, provenance predicate, SLSA level, or verification summary does not prove an uncited build/work/result relation, benchmark superiority, runtime safety, release approval, gate passage, or assurance. | Start with the attestation-bearing fields of that path slice, the produced-model/benchmark case, and CC-G6-07/08 when the adopted SLSA provenance/verification contract or in-toto Statement/v1 semantics change. |
| W3C Verifiable Credentials Data Model 2.0, Recommendation 15 May 2025 — adapt credential subject, issuer, holder, verifier, status, context, and validity separation for a path that cites an independently governed credential/status use. | PathCitationRecord separates source/carrier/currentness refs, downstream work, exact use relation, A.10 reliance disposition, and NotCarried; the dashboard-status case and CC-G6-09 require the status cue, query/use work, verifier or relying context, and actual reliance to remain distinct. | A valid credential, successful proof check, holder presentation, status value, or graph membership does not become claim truth, authorization, permission, gate passage, release, actual reliance, or assurance. | Start with those credential/status/use fields, the dashboard-status path, and CC-G6-09 if VC 2.0 or its adopted status/validity contract changes. |
| Pineau et al., Improving Reproducibility in Machine Learning Research, JMLR 22(164), 2021, and Mitchell et al., Model Cards for Model Reporting, FAT* 2019 — adapt exact method, dataset, metric, evaluation condition, version, limitation, and run-evidence disclosure as inputs to a replayable benchmark path. | The produced-model/benchmark case, dependency-closed PathSliceId, and CC-G6-02/07/08 keep model edition, training/evaluation work, dataset and method editions, local result, result episteme, source carrier, limitations, and currentness separately addressable. | A reproducibility checklist, model card, disclosed score, or limitation does not establish that training or evaluation occurred, that the reported result is current, that one model is superior, or that deployment is permitted. | Start with the model/benchmark slice fields, that worked case, and CC-G6-02/07/08 if the adopted reproducibility or reporting contract changes. |
| ISO/IEC/IEEE 15026-2:2022, Systems and software assurance — Part 2: Assurance case — adapt the separation between cited evidence and the structure, maintenance, and evaluation of an assurance case. | NotCarried names assurance explicitly, the subject-pattern map and §4.7 handle assurance under B.3, and CC-G6-10 permits the ledger to index evidence paths without becoming an assurance result. | A complete-looking evidence path, ledger entry, confidence label, or signed carrier is not an assurance claim, safety result, readiness result, compliance result, or release confidence. | Start with NotCarried, the B.3 extension boundary, one assurance-input path, and CC-G6-10 if the adopted assurance-case evidence or maintenance boundary changes. |
Start source return at the changed row’s named fields, rule, case and checklist rows. Then follow §4.5 through every actual dependent representation and receiving use, including another row when it shares the changed premise. Preserve independently supported accounts. The applicable subject patterns establish any revised relation or result before its graph projection is updated.
G.6:12 - Relations
- Builds on:
A.10for source recovery, provenance, bounded reliance and graph-edge discipline;A.10.1for source-change discovery and affected-use reach;A.2.4for first-use evidence/status classification;C.2.1for claim and result epistemes;C.29for representation correspondence. - Coordinates with:
A.13andA.15.1for the already identified performers and independently admitted Work;F.6andA.2.1only when the receiving path expressly relies on the claim that the Work was performed under a stated assignment;A.6.1for actual operation bindings; the exact pattern for each participation relation;A.15.PRODfor production or inception when current;C.16for measurement results;G.4for runtime evaluation results;B.1.6for work-resource aggregation results;C.28for causal use;F.10for status use;F.9for bridge and loss;E.18andE.18.2for transformation-flow structure;G.11for currentness;B.3for assurance;E.17for source-backed faces and return to source;E.24.PUBfor obtaining publication occurrences; and the pattern that defines each exact formal, diagnostic, conformance, comparison, selection, acceptance, gate, permission, commitment, or decision claim cited by a path. - Used by: selector, benchmark, replication, audit, refresh, assurance, maturity, and release patterns that need stable provenance-path citation, including
G.5,G.9, andG.11. - Does not govern: any represented work occurrence, participation, production, local result, result episteme, outcome, source publication, representation correspondence, currentness result, assurance, later use, or stronger conclusion named in
NotCarried.
G.6:End
G.7 - Cross‑Tradition Bridge Calibration Kit (BridgeMatrix → BridgeCards + BCT/Sentinels)
Status: Stable
Tag. Architectural pattern
Stage. design‑time (calibration + publication) + run‑time (sentinel‑driven telemetry emission; orchestration governed by G.11)
Start here. Name the correspondence to be calibrated and the question of the receiving use, if one is already known. Recover its exact endpoints and governing relation, then state the evidence, preserved distinctions and losses. The first useful result is a bounded calibration statement; its suitability for a receiving use requires that use’s own rule, loss tolerance and reliance. Add the kit fields below when calibrated values or published, refreshable calibration records are needed. For a bounded reuse needing neither, follow the correspondence’s direct rule instead (F.9 for senses, C.3.3 for kinds, or the named plane rule).
Primary output. A calibration kit with a BridgeCalibrationTable (BCT), CalibrationLedger, RegressionSet and SentinelSet. Each row cites the actual sense, kind or plane correspondence being calibrated. F.9 rows carry BridgeCards; public naming and real flow/gate crossings add their required UTS or bundle anchors. Sentinel triggers carry the affected references and live scope.
Working‑Model first. Prefer a minimal, auditable calibration procedure and worked micro‑cases; escalate to heavier harnesses only where risk warrants (per E.8).
G.7:1 - Problem frame
SoTA synthesis (G.2) can preserve pluralism by exporting a BridgeMatrix: a Tradition×Tradition inventory of “comparable constructs” with preliminary notes (candidate correspondences, likely losses, tentative levels). When the receiving use requires calibrated cross‑Context reuse, downstream patterns (CHR/CAL/selector/logging/shipping) must ensure that the reuse is:
- recoverable through the actual correspondence and its source basis, including BridgeCards for F.9 rows,
- calibrated with a small procedure whose evidence and conditions are inspectable,
- published with the anchors required by the receiving use; UTS and E.18/A.21 harnesses apply to their actual public names and flow/gate crossings,
- refreshable through the uses that depend on changed calibration, with path-scoped RSCR where paths are used.
The kit links BCT rows to their correspondence references, regression probes and sentinels for the uses that consume those rows. Later patterns can use it to satisfy the core invariants for cross‑Tradition reuse.
G.7:2 - Problem
- Cross-Tradition comparison through informal “synonymy” or an unspecified mapping can hide meaning drift and the relation needed for the comparison.
- An unexamined plane mismatch (world ↔ concept ↔ episteme, or another
ReferencePlaneshift) can support an unwarranted downstream confidence claim. - Calibration changes (CL/CL^k/plane or their policy pins) require rechecks of the uses that depend on them. Unrelated pack contents retain their existing basis.
- If bridges are involved in QD/illumination or other edition‑sensitive telemetry, edition pins must be tracked (otherwise comparisons become irreproducible after a map/distance/policy update).
- Averaging ordinal row values is inadmissible; even a lawful summary can hide a loss needed by a receiving task. Preserve the cell evidence and use only a summary justified for the stated calibration question.
G.7:3 - Forces
| Force | Tension |
|---|---|
| Comparability vs local authority | Enable comparisons across Traditions ↔ avoid overriding Context‑local meaning. |
| Auditability vs authoring throughput | Require explicit artefacts, losses, and pins ↔ keep the calibration procedure light enough to be used. |
| Targeted refresh vs safety | Emit scope-specific RSCR triggers ↔ retain the typed causes and payload needed for audit and planning further work. |
| Plane awareness vs “one story” | Explicitly surface ReferencePlane and plane penalties ↔ avoid turning plane discussion into a second semantics of “sameness”. |
| QD comparability vs metric drift | Enable cross‑context reporting of archive/illumination telemetry ↔ enforce edition‑aware pins for descriptor/distance/policies only when those modes are in use. |
G.7:4 - Solution — Bridge calibration kit (BCT + BridgeCards + RegressionSet/Sentinels)
G.7:4.1 - G.Core linkage (normative)
Builds on: G.Core (Part-G core invariants)
GCoreLinkageManifest (normative).
GCoreLinkageManifest := ⟨ CoreConformanceProfileIds := { GCoreConformanceProfileId.PartG.AuthoringBase, GCoreConformanceProfileId.PartG.TriStateGuard, GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted }, RSCRTriggerSetIds := { GCoreTriggerSetId.BridgeCalibrationKit }, CorePinSetIds := { GCorePinSetId.PartG.CrossingVisibilityPins }, CorePinsRequired := { BridgeCalibrationTableId (BCT.id), RegressionSetId, SentinelSetId, FreshnessWindowRef, CalibrationLedgerId, RowScopeId, ReferencePlane(src)?, ReferencePlane(tgt)?, UTSRowId[]?, PathId[]?/PathSliceId[]? }, DefaultsConsumed := ∅, TriggerAliasMapRef := ∅ ⟩
- Expansion rule. Expand the cited profile and set IDs, then take their union with the IDs listed above to obtain the effective
CoreConformanceIds,RSCRTriggerKindIdsandCorePinsRequired. Apply theG.Corenil‑elision and expansion rule. - Conditional pins.
- Relation, calibration and policy pins follow the channel/use conditions in G.Core:4.2.3. Source/target planes are required for an actual plane claim or another rule that consumes them; UTS and Path pins follow actual public-name and path uses.
BridgeCardRef.editionis required iff an F.9 BridgeCard is published as an editioned artefact.- Sentinel scopes MAY be recorded as
PatternScopeId[]when path surfaces are not available (and SHALL then be present in sentinel records and emitted trigger payload pins).
- CN/CG note.
CC‑GCORE‑CN‑CG‑1is included viaGCoreConformanceProfileId.PartG.AuthoringBaseand is exercised only when the governance card and legality gate (e.g.,CNSpecRef.edition/CGSpecRef.edition) are explicitly pinned; penalty/guard policy ids (Φ(CL),Ψ(CL^k),Φ_plane) are policy pins, not governance cards or legality gates.
(payload pins, minimum: affected members of the effective CorePinsRequired (after expansion) plus any pins introduced by active extensions (e.g., QD parity pins), scoped to the watched PathSliceId[]/PathId[]/PatternScopeId[].)
G.7:4.2 - Kit objects (surface governed by this pattern)
When a calibration kit is needed, G.7 defines its records, references and links for publication, regression checks and sentinel triggers. F.9 governs BridgeCard and CL. When the kind channel is used, C.3.3 governs KindBridge and CL^k.
(A) BridgeCalibrationTable (BCT) — object.
A BridgeCalibrationTable is a per‑Tradition‑pair registry of calibrated bridge entries.
Minimal fields (conceptual):
BridgeCalibrationTable := ⟨ BCT.id, TradPairId, FreshnessWindowRef, RowEntries[] ⟩
Provenance of G.2 rows. For a BCT derived from a G.2 BridgeMatrix, publish the BridgeMatrixId and the row references through G.7:Ext.MatrixIntake. Include BridgeMatrixRef.edition when the matrix is editioned. G.2 defines the source rows.
Where each RowEntry minimally binds:
RowEntry := ⟨ RowEntryId, ComparableConstructId, RowScopeId, BridgeCardId[]?, KindBridgeAssertionRef[]?, PlaneRelationRef[]?, RowCL_min?, RowCL_k_min?, RowCL_plane_min?, CalibrationBasisRef, LossNoteRef[]?, CounterExampleRef[]?, CounterExampleAbsenceRef?, ReceivingUseClaimRef[]?, ReceivingPolicyRef[]?, WaiverRef[]?, RegressionSetId, SentinelSetId, PolicyPins?: { Φ(CL)?, Ψ(CL^k)?, Φ_plane? }, PlanePins?: { ReferencePlane(src), ReferencePlane(tgt) }, ExtensionPins?: { [GPatternExtensionId]: { …ids… } } ⟩
(B) CalibrationLedger — object.
For each calibrated row, the CalibrationLedger identifies the calibration, its losses and the artefact or policy references a reader can check.
Minimal fields:
CalibrationLedger := ⟨ LedgerId, TradPairId, Entries[] // cite RowEntryId, relation/card refs, calibration basis and supported summaries, losses, counterexamples or search disclosure, UTS rows and any regression-run/delta refs; keep receiving-use/policy claims and a policy exception distinct ⟩
(C) RegressionSet — object.
A RegressionSet contains a small set of checks that can run against BCT entries. The checks can reveal calibration drift after changes in bridges, policies, planes or edition pins and supply evidence for RSCR triggers.
Minimal fields:
RegressionSet := ⟨ RegressionSetId, TradPairId, TestCaseId[], ExpectedOutcomesRef?, RegressionRunRef? ⟩
G.7:4.2.1 - Interpret calibration and assess a receiving use separately
Calibration question. State which correspondence, direction, scope, source editions and evidence the row assesses. Keep an F.9 sense correspondence, a C.3.3 kind correspondence and an applicable plane relation under their separate predicates. A record or favourable summary makes none of them obtain.
For a stated F.9 correspondence, optional CL shorthand means: 0 contradicted, 1 weakly comparable, 2 bounded support with explicit counterexamples, and 3 matched stated invariants with no current material counterexample. Cite the actual calibration basis. A kind-channel value follows C.3.3; a plane-channel value requires its own declared calibration rule. One channel cannot raise or replace another.
Summary meaning. Use RowCL_min only when a non-empty set of cells shares the declared ordinal scale and calibration question and the receiving report needs its weakest calibrated level. Cite that cell set and keep each loss recoverable. Apply the same condition independently to kind or plane summaries. An empty, mixed or unresolved basis has no such minimum; publish the separate results or the precise gap. A minimum is a calibration summary, not an admissibility result, and ordinal values are not averaged.
Counterexamples and search evidence. Preserve actual counterexamples and losses. Level 2 needs its cited bounded-support counterexample; level 0 identifies the contradiction. When evidence is weak or incomplete, disclose that limitation instead of reporting it as a discovered counterexample. At level 3, when none is cited, supply a citable search/absence account stating what was examined and that none was found or is currently known. This reports the search basis, not universal absence. A loss-noted row cannot be presented as free substitution.
Receiving use. For an F.9 use, state the separate claim with its use, direction, correspondence rule, loss tolerance and polarity. Recover the A.10 evidence-provenance path and direct-relation basis for that same bounded use. Apply the ordinary RelianceDisposition or, for an actual named assurance claim, the matching B.3 AssuranceResult under F.9:4.2, retaining the supported use, narrowing or stop that the applicable result warrants. A C.3.3 use separately checks receiving admissibility and target classification. Missing classification evidence remains unknown. Authorization, when required, follows its own rule. No CL level supplies these results.
A receiving policy may impose an additional threshold for one named use. Cite that use, the policy’s justification and authority, and its other necessary premises. The threshold is an additional policy condition, not F.9’s general law. A WaiverRef identifies an authorized exception to that policy only: it supplies no missing correspondence, target fact, suitable-use claim or evidence. Preserve any narrower allowed use and its independently supported conditions.
Plane and loss policies. When a receiving use relies on a plane relation, name the source and target planes, its predicate, calibration basis and applicable policy. Missing required plane information leaves that use unresolved. Keep any downstream abstain or policy-bound degrade under its receiving guard. Plane evidence does not rewrite CL or CL^k; a numerical loss requires its own receiving model and policy.
(D) SentinelSet & BridgeSentinel — object.
A SentinelSet identifies downstream uses to watch and links calibration changes to RSCR triggers for those uses.
Minimal fields:
BridgeSentinel := ⟨ SentinelId, watchedRowEntryIds: RowEntryId[], watchedRelationRefs: exact references to the correspondences assessed by those rows, watchedScope: PathSliceId[] | PathId[] | PatternScopeId[], payloadPins: { BCT.id, RegressionSetId, FreshnessWindowRef, affected RowEntryId[], watchedRelationRefs, PolicyPins?, PlanePins?, UTSRowId[]? } ⟩
SentinelSet := ⟨ SentinelSetId, BridgeSentinel[] ⟩
G.7:4.3 - Minimal calibration procedure
For each correspondence selected for calibration, grouped by Tradition pair and comparable construct:
- Recover the correspondence being calibrated. For an F.9 row, resolve the exact F.17 sense cells and profile, then produce or reuse its BridgeCard. A kind-channel row cites the C.3.3 kind endpoints and assertion; a plane row cites its own relation and rule. A coarser source label must be resolved to those actual endpoints before calibration.
- Record row scope and losses. Assign a
RowScopeIdand give each loss a citable reference, such as aLossNoteRef. Record the calibration basis and any meaningful channel summary under §4.2.1. If a receiving use is named, cite its separate suitability/classification and reliance results. Cite a waiver only for its exact authorized policy exception; it does not repair missing evidence. - Resolve any plane claim. If the row or receiving use consumes a plane relation, record its exact reference, source and target planes, governing rule and any plane policy applied. A kind-only or sense-only row creates no plane claim.
- Identify the policies used. Record policy and model references for a named threshold, exception, numerical loss or assurance calculation. Calibration without such a receiving use needs no invented Φ/Ψ/Φ_plane policy. Apply the loss and model-qualified assurance conditions of
CC‑GCORE‑PEN‑1to the actual use. - Summarize only a common calibration basis. Use the minimum only under §4.2.1’s conditions for a non-empty cell set, shared scale, shared calibration question and a report that needs the weakest level. If those conditions do not hold, report the separate cell results or the unresolved basis. In either case retain each loss and counterexample.
- Regression and sentinel wiring. Create/update the
RegressionSetandSentinelSet. Any calibration change that can affect downstream audit (CL/CL^k/plane pins, relevant policy ids, edition pins for involved telemetry surfaces, freshness window) emits typed RSCR triggers (canonical ids; scope + payload pins). If the regression harness is run, record a citableRegressionRunRef(or equivalent run/delta reference) and attach it to the relevant ledger entries (pin‑first; no narrative-only deltas).
G.7:4.4 - Publication surfaces (UTS + GateCrossing harness)
A conformant G.7 publication:
- publishes the exact correspondence references for each row, including BridgeCards for F.9 rows and UTS identifiers when the naming/publication rule requires them,
- makes an independently governed E.18 crossing or A.21 gate checkable through its applicable harness, preserving lexical, lane and required-pin constraints,
- emits RSCR triggers using canonical
RSCRTriggerKindIdand attaches the minimum payload pins listed in §4.1. - keeps SCR/Evidence citations complete for the use they support: include the row locator, exact correspondence basis and
{BCT.id, RegressionSetId}, plus the policy/model pins consumed by the reliance or assurance claim. Representation follows G.6/SCR when that surface is used.
G.7:4.5 - Worked mini-examples (row scopes and loss notes)
These worked rows use illustrative calibration values and source scopes. Actual calibration needs its stated evidence. Each receiving use has a separate rule and loss tolerance.
-
Preference‑learning objective (Method; RowScope = “training‑objective‑intent”). Cells:
RLHF@Context‑A↔DPO@Context‑B↔IPO@Context‑CRowCL_min: 2 (calibrated bounded support in this worked case) Loss notes: different inductive biases (reward model vs direct preference likelihood; sensitivity to preference noise model; implicit regularisation forms). Proposed use: a didactic comparison of objective intent. Its separate claim must limit the comparison to that intent and retain the listed differences; method eligibility and acceptance require their own rule. -
Robustness evaluation (Measurement; RowScope = “metric‑family‑intent”). Cells:
Accuracy@IID↔Robustness@ShiftBench(for example, a declared distribution-shift benchmark) RowCL_min: 2 Loss notes: shift taxonomy differs; comparability depends on pinned protocol editions and window selection; “robustness” is not a scalar substitute for accuracy. -
Quality‑Diversity archive comparability (Measurement; RowScope = “DescriptorMap‑only”). Cells:
MAP‑Elites grid indices↔CVT‑MAP‑Elites centroids↔CMA‑ME archiveRowCL_min: 2 Loss notes: discretisation vs centroidal tessellation; archive pressure differs; drift occurs ifDistanceDefor insertion policy changes. Proposed use: cross-reporting only the named descriptor-map relation under explicit edition pins and an affirmative bounded-use claim with passing reliance. Edition pins alone do not make the telemetry comparable. -
Open‑ended transfer semantics (Method; RowScope = “transfer‑rule intent”). Cells:
POET-class transfer rule↔Enhanced-POET-class transfer ruleRowCL_min: 2 Loss notes: environment validity region differs; transfer timing and selection pressures differ; pinning transfer rule editions is mandatory for audit.
Paired receiving case — Vehicle to TransportUnit. Use C.3.3 §9.1’s exact source and target kind declarations, pinned scheme editions, registryAPI v1.4 and selected time window. In row VehicleTransportOrder, record the obtaining KindBridge, preserved PassengerCar/Vehicle subkind order and collapsed EV distinction, with the reported CL^k=2 and battery-health loss. This is the kind channel; an F.9 sense Bridge is added only if the receiving claim separately relies on one.
For a G.5 shortlist of independently admitted Methods for a transport review, suppose the applicability criterion uses only the preserved transport/passenger order and explicitly ignores propulsion. The row can support that narrow applicability comparison after receiving admissibility, fresh target classification of the subject vehicles and the matching evidence-reliance result pass. The Methods’ other eligibility criteria remain applicable.
For a battery-health review whose Method-selection rule needs EV/battery information, the same correspondence fails that use because the required distinction is lost. A favourable CL value or a waiver cannot supply the battery premise. The correspondence and calibration can stay unchanged while these two use conclusions differ.
Use these two questions as a paired RegressionSet probe when this row is reused: recover the transport-order premise for the first and expose the missing battery premise for the second. Recheck the affected use after its criterion or the row’s preservation/loss basis changes.
In C.3.3 §9.3’s AdultPatient/AdultPerson_Y case, the age-boundary loss and CL^k=1 remain evidence about the kind correspondence. An authorized policy exception does not supply an unresolved date of birth; the receiving classification stays unknown.
Choose pins for the actual use. These cases apply the same conditions to a compact result and to the fuller kit:
| Use | What the result must retain |
|---|---|
| VehicleTransportOrder kind-only calibration | its C.3.3 kind endpoints, assertion, CL^k calibration basis, battery-health loss, row/freshness and kit references; a live sentinel can use PatternScopeId. No sense Bridge, plane relation or loss policy follows from this row. |
| F.9 sense-only calibration | exact F.17 sense endpoints, obtaining Bridge, BridgeCard and any reported CL basis; add neither a kind correspondence nor a plane claim without its own basis. |
| Plane-only calibration | the independently governed plane relation, planes and calibration rule/basis; add a numerical loss policy only if that receiving model is used. Plane change alone supplies no F.9 Bridge. |
| A G.2 harvest with no crossing, or a suite contract reused on another entity of the same kind | ordinary source/edition and applicability information; no crossing pin set is instantiated merely from the harvest, entity change or a new declaration edition. |
| A use relying on both a sense and kind correspondence | both independently established relations and their receiving conditions; neither channel replaces the other. |
| A safety-assurance comparison using an F.9 calibration, a defined plane-loss model and an A.21 gate | the exact Bridge/Card and row, BCT/regression/freshness evidence, actual plane relation and model/policy pins, matching A.10/B.3 reliance and assurance grounds, and every required gate anchor. A required missing pin leaves that use unresolved; favourable calibration alone grants no permission. |
G.7:4.6 - Extensions (pattern‑scoped; non‑core)
Extensions carry wiring only (pins/editions/policy‑ids + which governing patterns are applied). They MUST NOT redefine core invariants or defaults.
GPatternExtension: MatrixIntake
-
PatternScopeId:
G.7:Ext.MatrixIntake -
GPatternExtensionId:
MatrixIntake -
GPatternExtensionKind:
InteropSpecific -
GoverningPatternId:
G.2(BridgeMatrix semantics and comparable-construct inventory) -
Uses:
{G.2, F.9} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum):
BridgeMatrixId(and, if editioned:BridgeMatrixRef.edition)BridgeMatrixRowRef[](row‑level anchors for intake; defined by the governing pattern; e.g.,PatternScopeId/UTSRowId/ row ids)ComparableConstructId[](row keys; if the source does not supply a stable id,G.7mints one while preservingBridgeMatrixRowRefas the provenance anchor)LossNoteRef[]?(if exported byG.2; otherwise authored inG.7and cited from theCalibrationLedger)
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.EditionPinChange} -
Notes (wiring‑only): This module binds “row candidates” from G.2 to the BCT/Ledger intake without copying G.2 semantics into G.7.
GPatternExtension: DHCAccounting
-
PatternScopeId:
G.7:Ext.DHCAccounting -
GPatternExtensionId:
DHCAccounting -
GPatternExtensionKind:
DisciplineSpecific -
GoverningPatternId:
C.21(DHC metric semantics, including AlignmentDensity) -
Uses:
{C.21} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum; conditional on use):
AlignmentDensityMethodRef.edition?DeclaredUnitsRef?(the C.21 Unit for the reported quantity; AlignmentDensity usesobtaining_relations/100_compared_cells)
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.EditionPinChange} -
Notes (wiring‑only):
- G.7 records the counts and their declared units. C.21 defines what the counts mean and which operations are admissible.
- When reporting AlignmentDensity, follow C.21’s declared F.17 cell set and count only exact obtaining directed F.9 relations. Preserve each counted relation’s orientation and admitted-use qualifier, and keep observed loss in its evidence account. CL values neither change that count definition nor grant substitution;
CC‑G7‑DHC‑Units‑1checks the report’s units and cited method.
GPatternExtension: QDParityPins
-
PatternScopeId:
G.7:Ext.QDParityPins -
GPatternExtensionId:
QDParityPins -
GPatternExtensionKind:
InteropSpecific -
GoverningPatternId:
C.18(QD artefact semantics; uses C.19 for exploration/logging pins as needed) -
Uses:
{C.18, C.19} -
⊑/⊑⁺:
∅ -
RequiredPins/EditionPins/PolicyPins (minimum; conditional on use):
DescriptorMapRef.editionDistanceDefRef.editionInsertionPolicyRef(policy id or pinned policy ref, per governing definition semantics)
-
RSCRTriggerKindIds:
{RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent} -
Notes (wiring‑only): These pins identify the editions and policies needed to reproduce cross‑Context archive or illumination comparisons. C.18 governs the QD artefacts and C.19 supplies exploration or logging rules when needed. The pins from this module should be attached via
RowEntry.ExtensionPins[QDParityPins](or an equivalent extension‑pin map) and included inBridgeSentinel.payloadPinswhenever the watched scope consumes QD telemetry.
GPatternExtension: SoSLogClauses
- PatternScopeId:
G.7:Ext.SoSLogClauses - GPatternExtensionId:
SoSLogClauses - GPatternExtensionKind:
InteropSpecific - GoverningPatternId:
C.23(SoS‑LOG rule and branch semantics) - Uses:
{C.23, G.6} - ⊑/⊑⁺:
∅ - RequiredPins/EditionPins/PolicyPins (minimum; conditional on use):
SoSLogRuleId[](or governing definition‑equivalent ids)FailureBehaviorPolicyId?(policy id, when degrade behavior is bound)PathId/PathSliceIdcitations for explainability (viaG.6)BridgeCardId[](bridges whose reuse is being justified)
- RSCRTriggerKindIds:
{RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.MaturityRungChange} - Notes (wiring‑only): To justify a cross‑Tradition reuse decision, cite the applicable SoS‑LOG clauses and evidence paths. C.23 defines the clauses.
GPatternExtension: AcceptanceHooks
- PatternScopeId:
G.7:Ext.AcceptanceHooks - GPatternExtensionId:
AcceptanceHooks - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
G.4(Acceptance/threshold/unknown handling; G.7 does not define thresholds) - Uses:
{G.4} - ⊑/⊑⁺:
∅ - RequiredPins/EditionPins/PolicyPins (minimum; conditional on use):
AcceptanceClauseId[](or governing definition‑equivalent ids)AcceptancePolicyId?(policy id when acceptance behavior is pinned)BridgeCardId[](bridges whose calibrated status is being used as a gate input)
- RSCRTriggerKindIds:
{RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.BaselineBindingEdit, RSCRTriggerKindId.LegalitySurfaceEdit} - Notes (wiring‑only): When bridges are used as selector gates, thresholds and unknown-handling remain governed by Acceptance; this module only pins the linkage and refresh relevance.
GPatternExtension: AdvancedCalibrationProcedures
- PatternScopeId:
G.7:Ext.AdvancedCalibrationProcedures - GPatternExtensionId:
AdvancedCalibrationProcedures - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
G.7(calibration-result interface) - RequiredPins/EditionPins/PolicyPins: the selected domain calibration method and edition, its required inputs, and any policy used in the returned calibration.
- RSCRTriggerKindIds:
{RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.PenaltyPolicyEdit, RSCRTriggerKindId.ReferencePlaneEdit} - Use. When the minimal procedure cannot establish the calibration needed by the receiving question, select a domain procedure that supplies that inference, such as an uncertainty estimate under its stated model. Apply its prerequisites and retain the returned limitations. If no such procedure or required input is available, the needed calibration remains unestablished.
G.7:5 - Archetypal Grounding (System / Episteme)
System case: Cross-standard comparison of safety claims about one physical system. A team must compare a safety assurance claim across two regulatory Traditions (e.g., a “functional safety case” tradition and an “ML system testing” tradition) for the same physical system scope. Where “hazard”, “evidence carrier” or “pass criterion” has distinct source-local meanings, the team establishes the corresponding F.9 relations and uses G.7 to calibrate them. It records losses and any plane relation used in the comparison, then watches the affected uses so a protocol-edition change reopens their dependent claims under G.11.
Episteme case: Benchmark protocol pluralism. A research group wants to compare “state‑of‑the‑art” across multiple evaluation Traditions (IID performance, shift robustness, preference‑based evaluation). For each actual correspondence needed by that comparison, G.7 records the calibrated scope, preserved distinctions and losses, with BridgeCards for F.9 rows. Protocol-edition pins and sentinels identify affected selector uses when a protocol or policy changes; G.11 governs their reconsideration.
G.7:6 - Bias‑Annotation
A favourable calibration summary can hide a missing distinction required by the receiving task. Keep the calibration basis and that task’s premises separate, as in §4.5’s paired transport and battery-health uses.
G.7:7 - Conformance Checklist (normative) — CC‑G7
| ConformanceId | Requirement | Purpose |
|---|---|---|
| CC‑G7‑CoreRef | G.7 is conformant only if it satisfies the effective G.Core obligations declared by the GCoreLinkageManifest in §4.1 (after nil‑elision and expansion of profile/set/pinset ids), including any explicit deltas listed there. | Use the G.Core definition of each shared invariant by reference. |
| CC‑G7‑BCT‑1 | An active calibration kit has a BCT with its freshness basis, exact row scope, relation/card references, calibration basis and applicable evidence, regression and sentinel references. Channel summaries and policy/plane pins are present only under their declared calibration or receiving-use conditions. | Make the calibration and its use independently recoverable. |
| CC‑G7‑BridgeCard‑1 | An F.9 BridgeCard resolves its exact F.17 endpoint senses and relation profile. A kind-channel assertion follows C.3.3 and cites its exact kind endpoints; a plane claim cites its own governor. Keep the channels distinct. | Preserve the subject of each correspondence. |
| CC‑G7‑UTS‑1 | When G.7 mints or publishes a public identifier, apply CC-GCORE-UTS-1 and expose the resulting UTS rows in the consuming BCT/Ledger or crossing bundle. Include a BridgeCard or GateCrossing row only when that actual object is used; a kind-only calibration creates no F.9 BridgeCard. | Make public names citable without inventing relations. |
| CC‑G7‑RowScope‑1 | Every BCT row MUST declare its RowScopeId (which correspondence or difference is being calibrated), and any loss notes MUST be recorded as citable artefacts (refs/ids), not only narrative text. | Show which correspondence a later use relies on and what it can lose. |
| CC‑G7‑CLRegime‑1 | Every reported CL summary satisfies §4.2.1’s calibration meaning and aggregation conditions. Actual losses, counterexamples and required search/absence disclosures remain citable. The receiving use has its own rule, tolerance and reliance; a threshold or waiver cites its policy use, justification, authority and additional premises. | Calibration evidence supplies no automatic permission or receiving classification. |
| CC‑G7‑SCRLinkage‑1 | A calibration cited in SCR/Evidence surfaces MUST identify its exact sense, kind or plane correspondence, row locator and {BCT.id, RegressionSetId}. Add BridgeCard/UTS anchors for the actual F.9/public-name use and policy/model pins for the actual reliance, numerical loss or assurance calculation. | Preserve the evidence consumed by the claim without fabricating another channel. |
| CC‑G7‑SoSLOG‑Pins‑1 | When G.7:Ext.SoSLogClauses is in use, G.7 outputs MUST expose the cited SoS‑LOG rule ids and the relevant PathId/PathSliceId evidence citations; any change in those pins MUST be RSCR‑relevant per CC‑GCORE‑TRIG‑1…TRIG‑4. | Keep cross‑Tradition reuse explainable without embedding C.23 semantics. |
| CC‑G7‑Acceptance‑1 | When G.7:Ext.AcceptanceHooks is in use, G.7 outputs MUST expose the Acceptance clause ids/policy ids used as gates; thresholds/unknown handling remain governed by Acceptance; any change MUST be RSCR‑relevant per CC‑GCORE‑TRIG‑1…TRIG‑4. | Keep thresholds and unknowns out of bridges while preserving auditability. |
| CC‑G7‑RowBottleneck‑1 | Use a minimum only when the report needs the weakest level of a non-empty cell set sharing the declared ordinal scale and calibration question. If the conditions for the minimum do not hold, report separate results or the gap. Keep the constituent evidence and losses in either case. | Avoid averaging ordinal evidence or treating its minimum as a use decision. |
| CC‑G7‑PolicyPins‑1 | G.7 outputs MUST publish the policy id pins required to audit penalty routing and plane effects (ids only), as required by CC‑GCORE‑LINK‑1/2 and CC‑GCORE‑PEN‑1. G.7 MUST NOT duplicate policy tables or redefine penalty semantics. | Keep penalty routing auditable while preserving single‑governing-pattern policy semantics. |
| CC‑G7‑GateCrossing‑1 | An independently governed E.18 flow crossing or A.21 gate that consumes calibration MUST retain its required harness, pins, lexical constraints and lane checks. A calibration relation alone is not that flow crossing or gate. | Make actual crossings and gates checkable. |
| CC‑G7‑Sentinels‑1 | G.7 MUST register BridgeSentinel entries for bridges used by live scopes and MUST emit typed RSCR triggers (canonical RSCRTriggerKindId; see CC‑GCORE‑TRIG‑1…TRIG‑4) on calibration‑relevant edits, scoped to the watched PathSliceId[], PathId[] or PatternScopeId[], with the minimum payload pins from §4.1. | Identify the uses affected by a calibration change. |
| CC‑G7‑QD‑Pins‑1 | When G.7:Ext.QDParityPins is in use, G.7 outputs MUST include {DescriptorMapRef.edition, DistanceDefRef.edition, InsertionPolicyRef} and treat any change to those pins as RSCR‑relevant per CC‑GCORE‑TRIG‑1…TRIG‑4. | Prevent silent QD telemetry drift. |
| CC‑G7‑DHC‑Units‑1 | When AlignmentDensity is reported, G.7 outputs MUST count the exact obtaining directed F.9 relations in the declared F.17 cell set under C.21, include the declared units, and cite the active DHC method and replay basis. Related DHC accounts use their own exact C.21 definitions. CL labels grant no substitution and do not redefine the counted relation set; G.7 MUST NOT invent arithmetic over ordinal or otherwise inadmissible surfaces. | Keep dashboards and discipline-health readings faithful to their measurement definitions and bounded-use claims. |
G.7:8 - Common Anti-Patterns and How to Avoid Them
- Bridge‑by‑prose (“they have the same sense”). Avoid: establish the exact sense relation and its endpoints, then cite its calibration basis and row scope. Publish BridgeCards and UTS rows under their applicable kit and naming rules.
- Scope or sense relation used as a kind bridge.
Avoid: state the channel in
RowScopeIdand use its direct governor. A C.3.3 correspondence needs its kind endpoints and separate receiving classification; cite CL^k and any loss policy only when that calibration or reliance account uses them. - Plane blindness (“concept = world”). Avoid: identify the actual plane relation, its rule and evidence; record policy pins when that use consumes them, keeping plane effects separate from CL/CL^k semantics.
- CL smoothing / averaging. Avoid: establish the common scale and calibration question before taking a minimum; retain the counterexamples and losses, and keep the receiving decision separate.
- Pack‑wide refresh on a local bridge edit. Avoid: register sentinels for the watched path or pattern scopes and emit typed RSCR triggers with the changed basis and affected-use scope. G.11 follows the actual dependencies.
- QD metric drift by unpinned artefacts.
Avoid: enable
G.7:Ext.QDParityPinsonly when needed and require edition/policy pins when enabled.
G.7:9 - Consequences
- Auditable pluralism. Cross‑Tradition reuse becomes explicit, loss‑aware, and checkable.
- Targeted, edition-aware refresh. Calibration drift identifies the dependent uses to reconsider, including path-scoped RSCR where paths are used.
- Reuse in selection, logging and shipping. Each can cite the bridges and policy references it uses. They can use the G.Core crossing rules without defining local alternatives.
G.7:10 - Rationale
- Why publish a kit? Several consumers can use the same calibration. They use
CN‑Specfor the governing comparison rules andCG‑Specfor the legality gate. The kit supplies calibration without creating a second governance card or gate. - Why BCT + RegressionSet + SentinelSet? Regression tests make calibration drift detectable; sentinels identify the downstream scopes to refresh.
- Why declare the row scope? Two constructs may be comparable for one question and fail another. The scope identifies the correspondence or difference the row calibrates, so a reader can check the intended use.
G.7:11 - SoTA-Echoing — calibrate a correspondence for its receiving use
Practice question. When a correspondence has a favourable calibration, what must an engineer check before reusing it for a different task? The adopted line for this question keeps the exact relation, its justification and source versions recoverable, then tests the receiving task’s required distinctions. A benchmark-based alternative selects an alignment by its reference-benchmark score and carries that favourable score into downstream use.
SSSOM 1.0’s mapping model supplies the first line’s concrete separation of endpoints, relation predicate and justification. Its mapping FAQ distinguishes relation precision from confidence and explains why source versions and justification matter. Adopt that separation for the BCT row’s exact correspondence and calibration basis in §4.2. These sources address exchangeable ontology mappings; they do not establish an F.9 Bridge, a C.3.3 classification or permission for an FPF receiving use.
The OAEI 2025 Conference evaluation provides the benchmark comparator: its precision/recall measures assess generated correspondences against stated reference alignments and populations. That is useful for choosing a matcher on that question. Reject carrying its aggregate success, or one CL value, as a substitute for a different task’s premises. This is a limit of the inference, not a claim that OAEI promises such transfer.
Adapt the mapping-and-justification line in §4.2.1 and procedure steps 2/6: retain direction, scope and losses, and use a small regression pair whose required distinction changes. In §4.5, VehicleTransportOrder can retain passenger/transport order while losing battery information. The same calibrated relation can therefore support the stated transport comparison and fail the battery-health question. Keeping that distinction prevents a concrete erroneous use that a single favourable summary cannot detect.
At comparable effort, both alternatives start with the same relation row and existing calibration evidence. The selected line adds one named use condition and the smallest counterexample or regression pair that can change the answer. That is extra effort accepted for reuse across task boundaries; ordinary one-off correspondence handling remains with its direct sense, kind or plane rule when a calibration kit adds no value. SSSOM and OAEI provide the compared technical approaches, not empirical validation of this kit. Reopen when a new receiver needs a distinction outside the row’s stated preservation/loss basis, or when a cheaper rule can distinguish those same allowed and failed uses with equivalent support.
G.7:12 - Relations
Builds on: G.Core, G.2, F.3, F.7, F.9, F.17, B.3, E.10, E.18, A.21, G.6, C.21; C.3.3 when the kind channel is used.
Optionally uses via Extensions: G.4 (Acceptance hooks), C.23 (SoS‑LOG clauses), C.18 and C.19 (QD/OEE pins).
Used by / prerequisite for: In uses that consume this kit’s calibration, G.5 (cross‑Tradition eligibility/selection), G.11 (refresh orchestration), G.9 (parity across Traditions where bridges are required), G.10 (shipping surfaces that must cite bridge calibration ids), G.12 (DHC dashboards when bridge counts/units are surfaced).
Publishes to: UTS (bridge and crossing rows; twin labels as applicable) and emits RSCR‑ready telemetry/trigger payloads for G.11.
Constrains: Any downstream consumer that relies on this kit’s calibration claims or records must use the corresponding calibrated bridge artefacts/pins surfaced by this kit (governing G.Core crossing invariants apply).
G.7:End
G.8 - Package Method-Family Admissibility Rules and Maturity Ladders (SoS-LOG)
Status: Stable
Use this when. A selector needs a stable, citable package of C.23 admissibility rules and a maturity ladder for an already grounded G.5 method-family row. Resolve that row, the applicable rule editions and the evidence for the stated admission use, then publish their references as a bundle and maturity card. Later applications record their outcomes in the ledger.
Boundary. Use C.23 to define the rule itself. An ordinary G.5 result over grounded rows needs no G.8 bundle unless this packaged rule or ladder is consumed.
Policy references. The kit may cite the IDs of policies such as Φ/Ψ/Φ_plane, FailureBehaviorPolicyId, illumination promotion and E/E‑LOG. Cite the policies by ID only; their definitions stay with their governing sources. Under F.8:8.1, each must resolve, and a newly minted ID needs a resolvable PolicySpecRef and MintDecisionRef. G.8 MUST NOT silently mint policy IDs.
G.8:1 - Problem frame
Method families compete within a CG‑Frame. For lawful dispatch, a selector needs auditable tri-state admissibility decisions and evidence it can cite by ID. Cite any crossings the use depends on, and preserve alternatives in the set returned under a partial order. Rules spread across prose, dashboards and checklists can leave it without the rule edition, threshold source or evidence basis needed for its result.
The selector-facing, UTS-citable bundle identifies the applicable C.23 rules and cites the card for the method family’s ordinal or partially ordered maturity ladder. It connects these to Acceptance (G.4), evidence paths (G.6), selection and the registry (G.5), and refresh (G.11). Those patterns retain their governing definitions.
G.8:2 - Problem
- Selector needs a stable input artefact. Maturity narratives and scattered SoS‑LOG excerpts leave
G.5to reconstruct the rules or guess their defaults. - Thresholds leak into LOG. Numeric gates embedded in rule text or ladder rungs blur the boundary between LOG decisions (
C.23) and Acceptance thresholds (G.4). - Auditability is brittle. Decisions (
pass/degrade/abstain) lack stable, citable links to evidence paths (G.6) and crossing pins (Bridge/CL/Φ policy ids as required for the stated use), so later re‑checks and RSCR become ad‑hoc. - Telemetry contaminates decision semantics. Undeclared use of QD/OEE/illumination signals in dominance, or a changed definition edition, can alter outcomes without making the new basis visible.
- Refresh is under‑specified. Bundle evolution (rules, ladders, pins, policies, editions) must be RSCR‑addressable via typed trigger kinds, not by free‑text “reasons”.
G.8:3 - Forces
| Force | Tension |
|---|---|
| Pluralism vs. dispatchability | Preserve multiple method families and partial orders ↔ provide a consumable artefact for G.5. |
| Auditability vs. authoring friction | Fine‑grained pins and citations ↔ keeping authoring lightweight and notation‑independent. |
| Maturity as poset vs. scalar ranking | The declared ladder may be ordinal or partially ordered ↔ a single readiness number can hide its distinctions. |
| Telemetry richness vs. decision hygiene | Rich QD/OEE telemetry ↔ avoid illegitimate promotion into dominance without explicit policy. |
| Design‑time packaging vs. run‑time trace | Authoring produces stable bundles ↔ run‑time produces branch‑specific path traces and admissibility ledgers. |
| Interoperability vs. crossing discipline | Reuse across meanings, kinds or planes ↔ establish the actual correspondence and its receiving-use conditions. |
G.8:4 - Solution — Publish SoS‑LOG bundles and maturity cards as UTS‑citable kit
G.8:4.1 - G.Core linkage (normative)
Builds on: G.Core (Part-G core invariants)
GCoreLinkageManifest (normative).
(Canonical shape, Nil‑elision, and Expansion rule are per G.Core:4.2.)
Method- and generator-specific pins. Specify method- or generator-specific pins, editions and policies only in the GPatternExtension modules of G.8:5. The bundle and ledger schemas may list these as optional fields activated by the relevant extension. The core manifest lists only base-kit pins and Part-G-wide links.
GCoreLinkageManifest := ⟨
CoreConformanceProfileIds := {
GCoreConformanceProfileId.PartG.AuthoringBase,
GCoreConformanceProfileId.PartG.TriStateGuard,
GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted,
GCoreConformanceProfileId.PartG.ShippingBoundary
},
RSCRTriggerSetIds := { GCoreTriggerSetId.EvidenceGraphKit },
CorePinSetIds := {
GCorePinSetId.PartG.AuthoringMinimal,
},
CorePinsRequired := {
// Public ids governed by this pattern (strengthen conditional pins where G.8 publishes UTS publication units)
UTSRowId[], // bundle/ledger/card rows + any referenced UTS rows
SoS‑LOGBundleRef,
SoSLogRuleId[],
MethodFamilyRowRef, // exact G.5 <MethodFamilyId, rowEdition>
RegistrationContext,
// Closed value sets (ids only; UTS-registered)
DegradeModeEnum,
MaturityRungs,
// Maturity ladder pins
MaturityCardRef, // required; recommended: published as separate UTS artefact
MaturityRungId?, // iff a specific rung is asserted at packaging/run-time
// Evidence / provenance pins
A10EvidenceGraphRef?[], // packaging-time A.10 carriers (when PathId/PathSliceId not yet available)
EvidenceGraphId?, // iff resolvable to G.6 EvidenceGraph
PathId[]/PathSliceId[]?, // run-time ledgers typically have them
// Authoring traceability (SoTA-of-description)
AuthoringMethodDescriptionRefs?[], // edition-pinned method-description refs
},
DefaultsConsumed := {
DefaultId.PortfolioMode,
DefaultId.DominanceRegime,
DefaultId.GammaFoldForR_eff
},
⟩
(RSCR payload pins typically include: SoS‑LOGBundleRef, SoSLogRuleId[], MaturityRungId?, and EvidenceGraphId/PathId/PathSliceId?.
Crossing payload pins follow the actual sense, kind or plane correspondence and receiving use under G.Core:4.2.3; F.9 sense reuse uses G.8:Ext.BridgeReuseWiring. CL and loss-policy pins are required only by the actual calibration or separate named assurance use.
Method-/generator‑specific payload pins are listed only inside the relevant GPatternExtension blocks in G.8:5.)
Include DefaultId.GammaFoldForR_eff in DefaultsConsumed only if the bundle or ledger exports aggregated R_eff summaries. Otherwise apply nil-elision.
G.8:4.2 - Kit: objects and naming discipline (LEX heads; twin‑register safe)
Objects / surfaces (pattern-governed).
-
SoS‑LOG.RuleAn executable tri‑state decision schema{pass | degrade(mode) | abstain}for(TaskSignature, MethodFamily), cited by a rule id. (“pass” may be described as “admit” in prose, but the normative tri‑state vocabulary isG.Core’s{pass|degrade|abstain}.)C.23defines the rule.G.8packages its ID and the pins that bind it. -
SoS‑LOGBundle@ContextA selector‑facing, notation‑independent packaging object published to UTS. -
AdmissibilityLedger@ContextA run‑time ledger view that records admissibility outcomes, cited evidence paths, branch tokens, and the pins required for audit/refresh. -
MethodFamily.MaturityCardDescription@ContextA maturity ladder description published as a citable artefact: ordinal/poset, closed rungs,ReferencePlanedeclared; no thresholds inside.
Naming discipline (E.10 + “Spaces ≠ Maps”).
-
Technical heads are normative; Plain twins are didactic only and MUST NOT cross kinds.
-
Do not alias
CharacteristicSpaceandDescriptorMap.DescriptorMapRefis a map‑reference (typically used with QD archives).CharacteristicSpaceRefis a space‑reference (grid/cell semantics, if used).
-
A
CharacteristicSpaceRefidentifies the exact space. When the use depends on its defining episteme, carrycharacteristicSpaceDescriptionRef: U.EpistemeRefseparately and resolve the pair under A.19:5.1.1.1. Other required edition pins follow their direct reference patterns; informal names supply no pin.
G.8:4.3 - SoS‑LOGBundle@Context schema (conceptual; notation‑independent)
A conforming bundle is published to UTS. Its field meanings are fixed, and its internal representation may vary:
SoS-LOGBundle@Context :=
⟨
UTS.id := SoS‑LOGBundleRef,
Edition,
// Scope + spec pins (from GCorePinSetId.PartG.AuthoringMinimal)
CG-FrameContext,
entityOfConcern, GroundingHolon?, ReferencePlane, // subject of the LOG claim, empirical ground when used, and plane
CNSpecRef.edition,
CGSpecRef.edition,
MethodFamilyRowRef, // exact G.5 <MethodFamilyId, rowEdition>
RegistrationContext,
SoSLogRuleId[] , // ids only; semantics governed by C.23
ClosedEnums: {DegradeModeEnum, MaturityRungs}, // ids only; UTS-registered closed value sets
A10EvidenceGraphRef?[] , // packaging-time evidence carriers (A.10 anchors) when paths are not yet stable
MaturityCardRef , // UTS ref to maturity card (required; may be embedded but MUST be citable)
MaturityRungId? , // if a specific rung is asserted at packaging time
// Optional: Acceptance wiring (thresholds remain governed by G.4)
AcceptanceClauseId[]? ,
// Optional: Evidence wiring (for later audit & rung transition justification)
EvidenceGraphId? ,
PathId[]/PathSliceId[]? ,
// Optional: correspondence wiring, under the actual channel/use conditions of G.Core:4.2.3
BridgeId/BridgeCardId? ,
CL/CL^k/CL^plane? ,
Φ/Ψ/Φ_plane policy-ids? ,
// Optional: selector semantics pins (explicit value or resolved via DefaultGoverningDefinitionIndex)
PortfolioMode? ,
DominanceRegime? ,
// Optional: QD / OEE pins (only when those surfaces are declared)
CharacteristicSpaceRef? ,
characteristicSpaceDescriptionRef? , // U.EpistemeRef; exact definition under A.19:5.1.1.1
DescriptorMapRef.edition? ,
DistanceDefRef.edition? ,
EmitterPolicyRef? ,
InsertionPolicyRef? ,
// Optional: Open-ended pins (only when those surfaces are declared)
GeneratorFamilyRowRef? , // exact G.5 <GeneratorFamilyId, rowEdition>
EnvironmentValidityRegionId? ,
CouplerPolicyId? ,
TransferRulesRef.edition? ,
// Optional: branch/failure wiring (policy-bound)
FailureBehaviorPolicyId? ,
SoSLogBranchId[]? ,
// Optional: authoring traceability (SoTA-of-description)
AuthoringMethodDescriptionRefs?[] ,
Notes
⟩
Bundle requirements
SoS‑LOGBundle@Contextdoes not introduce new legality or normalization rules; it cites the pinned references above.- Thresholds and numeric gates are cited by id from
G.4Acceptance (no embedding inside the bundle). - Cite the actual correspondence and receiving-use grounds under G.Core:4.2.3, with F.9 wiring through
G.8:Ext.BridgeReuseWiringwhen applicable. Evidence paths are citable when available.
Binding obligations B1–B5
- B1 — Evidence wiring. At packaging time the bundle SHOULD provide resolvable evidence refs (typically
A10EvidenceGraphRef?[]and/orEvidenceGraphId?). At run time, admissibility outcomes SHOULD citePathId/PathSliceIdwhen available (G.6), so rung transitions anddegrade/abstaintraces are audit‑stable. - B2 — CL/plane routing pins. When reuse depends on a sense, kind or plane correspondence, the bundle/ledger MUST cite its obtaining relation and receiving-use grounds under G.Core:4.2.3. An F.9 use keeps its separate bounded-use claim and reliance basis through
G.8:Ext.BridgeReuseWiring; kind and plane uses retain their direct governing conditions. The bundle/ledger MUST pin the Bridge/CL/Φ/Ψ/Φ_plane references required by that use (reference-only; policy ids resolvable perF.8:8.1). CL and loss-policy pins are mandatory only when required by the calibration or separate named assurance account. Any numerical assurance penalty MUST satisfyCC‑GCORE‑PEN‑1, including its model applicability and declared-result conditions. - B3 — Retention policy and QD fields. Record
PortfolioModeand dominance inputs under CC-G8-5 and their G.5 definitions. When the bundle or ledger binds a QD or QD-using generator branch, pin the descriptor, distance, insertion and emitter definitions required for that use under CC-G8-6.PortfolioMode=Archivealone preserves evidence and requires no QD definitions. Any illumination telemetry remains report-only unless aG.4governing-pattern policy authorises its dominance contribution and that policy id is recorded in the run-time trace. - B4 — Open-ended fields. If the bundle binds an open-ended generator family, pin
GeneratorFamilyRowRef,TransferRulesRef.editionand any validity-region/coupler policy ids used. For unknown transfer validity, apply C.23 R0/R2/R3/R7: use a declareddegradebranch only after its prerequisites pass and its narrower use is permitted; otherwise returnabstain. Retain the unknown condition and branch basis. - B5 — Telemetry hooks. On any material telemetry event (illumination increase, archive insertion, probe accounting update, open‑ended coverage/regret proxy update), the emitted telemetry pins SHOULD include the controlling policy ids plus the relevant edition pins (e.g.,
DescriptorMapRef.edition,DistanceDefRef.edition,TransferRulesRef.edition) and, when available,PathSliceIdto keep RSCR planning auditable.
G.8:4.4 - AdmissibilityLedger@Context (run‑time view; selector‑facing)
A conforming ledger is a UTS-published view (or a view-projection of a Work/Audit artefact) of the branch reached in the run. Its row positions are:
⟨ MethodFamilyRowRef, TaskSignatureRef, SoSLogRuleId, RuleEdition, EvidenceProfileRef?, ClaimScope?, QualificationWindow?, IntendedAdmissionUse, EligibilityVerdict?, CGSpecVerdict?, AcceptanceVerdictRefs?, PolicyEditionRefs[], GuardDecision ∈ {pass|degrade|abstain}, DegradeMode?/SoSLogBranchId[]?, MaturityRungId?, AcceptanceClauseId[]?, EvidencePathRefs?, CrossingPins?, PortfolioMode?, DominanceRegime?, Edition ⟩
The bundle, maturity card and ledger bind one exact G.5 MethodFamilyRowRef = <MethodFamilyId, rowEdition>; an open-ended generator use also binds GeneratorFamilyRowRef = <GeneratorFamilyId, rowEdition>. Their publication Edition does not replace a registry-row edition. Resolve an earlier result against the row edition it cites.
Retain every premise established or validly reused for the branch, with its applicable evidence profile, claim scope, qualification window, intended use and policy editions. Follow C.23:4.2.1 when a required reference is unavailable or evaluation stops early: name the missing basis or the judgement not reached, and leave its dependent result unestablished. Identify an applicable earlier result as reused.
These report qualifications add no guard status or maturity rung. pass requires the complete admission basis, and degrade requires the permitted branch and its prerequisites.
Use EvidencePathRefs to cite the relevant PathId[] or PathSliceId[] when the outcome’s evidence is addressable through a G.6 path or slice. CrossingPins retain the correspondence and receiving-use grounds required by G.Core:4.2.3. For G.8:Ext.BridgeReuseWiring, these include the F.9 bounded-use claim, reliance basis and applicable Bridge/CL/Φ pins.
G.8:4.5 - Maturity ladder as a citable poset (published card)
MethodFamily.MaturityCardDescription@Context names the exact MethodFamilyRowRef, evidence profile, claim scope and selected slices, qualification window, and intended admission use required by C.23. It is published with:
- closed rungs (UTS‑registered identifiers),
Scale kind = ordinaland a declaredReferencePlane,- the order needed for any comparison the admission use makes, stated here or recoverable from the cited scale edition,
- rung transition justifications that cite evidence paths (typically
G.6paths).
For an “at least” floor, check whether the assessed rung equals or succeeds the required rung in that order. Known incomparability fails that test. An unavailable comparison leaves the verdict undetermined under the clause’s unknown rule. A clause that calls for a particular kind of evidence also needs grounds establishing that fact.
Dispatchers, auditors and refresh planners can cite the published ladder description. The card describes the ladder under C.23.
G.8:4.6 - Interfaces (minimal I/O standard; conceptual)
| Interface | Consumes | Produces |
|---|---|---|
G.8‑1 Publish_LOGBundle | MethodFamilyRowRef, SoSLogRuleId[] with rule editions (C.23), pins to Acceptance/Evidence/Crossings (as applicable) | SoS‑LOGBundle@Context (UTS row) |
G.8‑2 Publish_AdmissibilityLedger | Bundle + run‑time branch outcomes + evidence path refs (when available) | AdmissibilityLedger@Context (UTS row or UTS‑citable view) |
G.8‑3 Publish_MaturityCard | Ladder description + (optional) evidence path refs for rung transitions | MaturityCardDescription@Context (UTS row; editioned) |
G.8‑4 Expose_TelemetryHooks | QD/OEE/archive/open‑ended telemetry signals (when declared) | telemetry pins for refresh (…Ref.edition, policy‑ids, PathSliceId when available) |
G.8:5 - Extensions (pattern‑scoped; non‑core)
Publish method- or generator-specific requirements in GPatternExtension blocks with stable PatternScopeIds. The core kit is shared across methods and generators.
G.8:5.1 - G.8:Ext.SoSLOGWiring
PatternScopeId: G.8:Ext.SoSLOGWiring
GPatternExtensionId: SoSLOGWiring
GPatternExtensionKind: MethodSpecific
GoverningPatternId: C.23
Uses: {C.23}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
SoSLogRuleId[]SoSLogBranchId[]?FailureBehaviorPolicyId?(when degrade behaviour is policy‑bound)
RSCRTriggerSetIds / RSCRTriggerKindIds: ∅ (covered by G.8:4.1)
Notes (wiring‑only):
- Rule meaning, branch taxonomy, and “probe/sandbox” semantics are governed by
C.23; this module only binds ids and pins.
G.8:5.2 - G.8:Ext.AcceptanceWiring
PatternScopeId: G.8:Ext.AcceptanceWiring
GPatternExtensionId: AcceptanceWiring
GPatternExtensionKind: MethodSpecific
GoverningPatternId: G.4
Uses: {G.4}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
AcceptanceClauseId[]EvidenceProfileId[]?(if the ledger/bundle cites evidence profile ids rather than only paths)PromotionPolicyId?(only if telemetry may be promoted into dominance by explicit CAL policy)
RSCRTriggerKindIds (optional delta): {RSCRTriggerKindId.PolicyPinChange} (only if acceptance policies are pinned as ids in the bundle/ledger)
Notes (wiring‑only):
- Thresholds remain governed by
G.4Acceptance; this module carries only clause ids and policy pins.
G.8:5.3 - G.8:Ext.BridgeReuseWiring
PatternScopeId: G.8:Ext.BridgeReuseWiring
GPatternExtensionId: BridgeReuseWiring
GPatternExtensionKind: InteropSpecific
GoverningPatternId: G.7
Uses: {G.7, F.9, A.10, B.3}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum; conditional on the stated use):
BridgeId/BridgeCardId(the obtaining Bridge used; aBridgeCardIdis needed only when that Card is relied on)CL/CL^k/CL^plane(when cited; the applicable values are mandatory when required by theG.7calibration or namedB.3assurance account)Φ/Ψ/Φ_plane policy-ids(only the policy ids and editions required by the actualG.7calibration or namedB.3assurance account; reference‑only and resolvable perF.8:8.1)BridgeCalibrationTableId?,RegressionSetId?(both required when calibration evidence is cited, together with the row locator and active policy pins required byG.7CC‑G7‑SCRLinkage‑1)
RSCRTriggerSetIds: {GCoreTriggerSetId.BridgeCalibrationKit} (only if the bundle/ledger explicitly binds calibration records by id)
Notes (wiring‑only):
- Present when
SoS-LOGBundle@Contextuses an F.9 sense correspondence. A combined kind or plane use retains its separate relation and receiving conditions under G.Core:4.2.3. - The wiring MUST keep the obtaining Bridge reference, the separate bounded-use claim (use, direction, rule, and tolerated loss), and the
A.10reliance basis citable. A separate named assurance use also cites its exact target claim, receiving use, andB.3assurance basis/result. Required CL values, policy editions, and evidence remain mandatory for that account; any numerical assurance penalty satisfies the model applicability and declared-result conditions ofCC‑GCORE‑PEN‑1. Ordinary supported reuse does not require a CL shorthand, calibration record, loss-policy id, or assurance claim merely to fill the package.
G.8:5.4 - G.8:Ext.QDArchiveTelemetry
PatternScopeId: G.8:Ext.QDArchiveTelemetry
GPatternExtensionId: QDArchiveTelemetry
GPatternExtensionKind: MethodSpecific
GoverningPatternId: C.18
Uses: {C.18, G.5}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
DescriptorMapRef.editionDistanceDefRef.editionEmitterPolicyRefInsertionPolicyRefCharacteristicSpaceRefwith separatecharacteristicSpaceDescriptionRef: U.EpistemeRef(required iff cell boundaries / de‑dup / parity depend on the space definition; resolve the pair under A.19:5.1.1.1)
RSCRTriggerKindIds: {RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange}
Notes (wiring‑only):
- Archive/illumination signals are telemetry; promotion into dominance is only via explicit
G.4policy pins.
G.8:5.5 - G.8:Ext.ExploreExploitTelemetry
PatternScopeId: G.8:Ext.ExploreExploitTelemetry
GPatternExtensionId: ExploreExploitTelemetry
GPatternExtensionKind: MethodSpecific
GoverningPatternId: C.19
Uses: {C.19}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
ExploreExploitBudgetPolicyId?ProbeAccountingId?
RSCRTriggerKindIds: {RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.PolicyPinChange}
Notes (wiring‑only):
- When “probe/sandbox” is used, the controlling policy ids are pinned and recorded in the ledger/bundle trace.
G.8:5.6 - G.8:Ext.OpenEndedWiring
PatternScopeId: G.8:Ext.OpenEndedWiring
GPatternExtensionId: OpenEndedWiring
GPatternExtensionKind: GeneratorSpecific
GoverningPatternId: G.5 (generator family registry surface; algorithm semantics remain external to Part‑G core)
Uses: {G.5}
⊑/⊑⁺: ∅
RequiredPins/EditionPins/PolicyPins (minimum):
GeneratorFamilyRowRefTransferRulesRef.editionEnvironmentValidityRegionId?CouplerPolicyId?
RSCRTriggerKindIds: {RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta}
Notes (wiring‑only):
- Open‑ended coverage/regret (or similar) remains telemetry unless explicitly promoted by a governing-pattern policy.
G.8:6 - Archetypal Grounding (System / Episteme)
Show‑A — Tri‑state admissibility with set‑valued selection (multi‑criteria).
A CG‑Frame carries multiple offline/robust decision families (for example, conservative offline RL and transformer-based policy models). The bundle cites SoSLogRuleId[] (SoS‑LOG semantics in C.23) and cites AcceptanceClauseId[] for any floors (governed by G.4). The run‑time AdmissibilityLedger cites PathSliceId (when available) for each pass/degrade/abstain. G.5 uses the ledger to return a selected set under the declared partial order. A scalar winner would discard alternatives that order leaves unresolved.
Show-B — QD archive dispatch with edition-pinned descriptors.
A method family uses a QD archive, for example one driven by CMA-ES or a differentiable QD method. The bundle pins DescriptorMapRef.edition and DistanceDefRef.edition, plus insertion/emitter policies. Illumination metrics are logged as telemetry; any promotion into dominance is only via explicit CAL policy pins (recorded in the admissibility trace).
Show-C — Open-ended environment–method co-evolution.
A generator family operates in an open-ended setting (e.g., POET-style or PAIRED-style regimes). The bundle carries TransferRulesRef.edition and validity-region pins. If transfer validity is unknown, a declared narrower-use branch may return degrade after C.23’s prerequisites pass; without that permitted branch, the result is abstain.
If the evidence profile is missing at R0, report that missing prerequisite. Record the later Acceptance judgement as not reached; assigning it false would invent a result. Retain an applicable maturity result when it was validly reused. Coverage/regret telemetry remains report-only unless its governing policy promotes it into dominance.
G.8:7 - Bias‑Annotation
A maturity label can hide different evidence or admission conditions. Keep the exact rule, row edition and receiving use beside the label; the bundle itself supplies no maturity verdict.
G.8:8 - Conformance Checklist (CC‑G8)
-
CC‑G8‑CoreRef (G.Core conformance bridge). A conforming
G.8SHALL satisfy the effective set ofCC‑GCORE‑*obligations implied byG.8:4.1(expanded perG.Core:4.2), including required pins, trigger sets, and Default Governing Definition Index citation. -
CC‑G8‑1 (No thresholds in LOG). Any numeric gate, maturity floor, or threshold SHALL be authored as a
G.4Acceptance artefact and cited by id; the LOG bundle/ladder SHALL NOT embed thresholds. -
CC‑G8‑2 (Tri‑state discipline; delegated). Guard outcomes SHALL obey the tri‑state domain and unknown handling defined in
G.Core(delegation toCC‑GCORE‑GUARD‑1). Any sandbox/probe‑only behaviour SHALL be represented as an explicitC.23branch and MUST pin (and record) the controlling policy id (typically an E/E‑LOG policy id viaC.19), rather than inventing a fourth status or silently coercing unknowns. -
CC‑G8‑3 (Path citation when evidence is path‑addressable). For each recorded
pass/degrade/abstainoutcome whose evidence is path-addressable underG.6, the run-timeAdmissibilityLedgerMUST cite thatPathId/PathSliceId. At packaging time, the bundle/ledger SHALL at minimum provide resolvable evidence refs (e.g.,EvidenceGraphId?+ anchor refs). -
CC‑G8‑4 (Crossing visibility and penalty routing; delegated). Any cross-Context or cross-plane reuse asserted by the bundle/ledger SHALL satisfy the core crossing visibility and penalty routing invariants (delegation to
CC‑GCORE‑CROSS‑1andCC‑GCORE‑PEN‑1). -
CC‑G8‑5 (PortfolioMode/dominance hygiene; delegated). The bundle/ledger SHALL treat
PortfolioModeand dominance fields as pinned inputs and SHALL cite the governing definition for each omitted default throughG.Core.DefaultGoverningDefinitionIndex(delegation toCC‑GCORE‑DEF‑1andCC‑GCORE‑SET‑1; governing definitions includeCC‑G5.23forDefaultId.PortfolioModeandCC‑G5.28forDefaultId.DominanceRegime). It MUST NOT restate default values locally. If the bundle/ledger records telemetry that could influence dispatch (e.g., illumination/QD/OEE/open‑ended proxies), such telemetry SHALL remain report‑only unless explicitly promoted by aG.4governing-pattern policy id that is pinned and recorded in the run‑time trace. -
CC‑G8‑6 (QD/OEE edition discipline). When QD/OEE surfaces are declared, the bundle/ledger MUST pin the relevant editions and policies (
DescriptorMapRef.edition,DistanceDefRef.edition, insertion/emitter policies, andTransferRulesRef.editionwhen applicable). The exactCharacteristicSpaceRefand separatecharacteristicSpaceDescriptionRef: U.EpistemeRefare required iff cell boundaries / de‑dup rules / parity depend on the space definition. Resolve the pair under A.19:5.1.1.1; neither substitutes forDescriptorMapRef.edition. -
CC‑G8‑7 (Maturity is ordinal/poset). Maturity ladders SHALL be authored as ordinal/poset descriptions with closed rung ids (
MaturityRungs, UTS‑registered) and a declaredReferencePlane, and SHALL be published as a citable UTS artefact (editioned; twin‑register safe). Rung transitions, when asserted, MUST be justifiable by citable evidence paths (when available). -
CC‑G8‑8 (Spaces ≠ Maps).
CharacteristicSpaceandDescriptorMapSHALL remain distinct kinds; naming and twin‑register discipline must be respected. -
CC‑G8‑9 (Notational independence). The bundle, ledger, and maturity card SHALL remain notation‑independent (per
E.5.2); any serialization choice is non‑normative and belongs outside Part‑G core. -
CC‑G8‑10 (MOO cross‑reference). When a LOG bundle is used to drive or justify a produced selected-set outcome, the record of the producing Work, or the audit artefact, SHOULD cite the controlling mechanism ids (e.g., parity/shipping/refresh artefact ids) and relevant policy pins.
-
CC‑G8‑11 (SoTA‑of‑description trace). If authoring methods (e.g., discovery, clustering, summarisation) materially shaped rule text or rung definitions, the bundle/card SHOULD cite their method description refs (edition‑pinned) to support cross‑stance traceability.
G.8:9 - Common Anti‑Patterns and How to Avoid Them
-
Anti‑pattern: Embedding thresholds inside SoS‑LOG rules or ladder rungs. Avoid: thresholds live in
G.4Acceptance; bundle only cites clause ids. -
Anti‑pattern: Treating illumination/QD telemetry as a hidden scalar score that changes dominance. Avoid: keep telemetry report‑only unless explicitly promoted by a governing-pattern policy pin.
-
Anti‑pattern: Publishing a bundle that implies a correspondence or receiving-use conclusion without its required grounds, or omits CL/Φ pins required by the actual calibration or named assurance use. Avoid: retain the actual channel and receiving-use grounds under G.Core:4.2.3, using
G.8:Ext.BridgeReuseWiringfor F.9 reuse. Missing required grounds block the dependent use. -
Anti‑pattern: Re‑defining
PortfolioMode/DominanceRegimedefaults in the bundle text. Avoid: cite each default’s governing definition throughG.Core.DefaultGoverningDefinitionIndex. -
Anti‑pattern: Recording RSCR “reasons” as prose labels only. Avoid: emit canonical
RSCRTriggerKindIdvalues perG.Core.
G.8:10 - Consequences
- Positive:
G.5receives a stable, citable, selector‑facing artefact without importing rule semantics or threshold logic. - Positive: Pins, crossing references, evidence paths and trigger kinds tell an auditor what grounded the decision and help identify the uses whose checks may need to be reopened.
- Positive: The ladder retains its ordinal or partial-order meaning, so a maturity label cannot substitute for the evidence and conditions of admission.
- Negative: Requires stricter authoring discipline (UTS publication, pin completeness, explicit wiring).
- Negative: Without maintained G.6 paths, a later reader may need more work to recover the cited evidence. Evaluate the resolvable A.10 anchors and their support for the stated use; absence of a graph alone does not lower evidence-support class. If a required basis cannot be recovered, abstain from the dependent use.
G.8:11 - Rationale
C.23 defines the rules, G.4 defines Acceptance and its thresholds, G.6 supplies path-addressable provenance, and G.5 defines selection and registry use. A project can cite all four in one auditable package for dispatch without maintaining competing definitions.
G.8:12 - SoTA-Echoing
These examples illustrate how applications connect decision rules, acceptance thresholds, provenance and set-valued outputs under their respective governing rules:
- Set-valued selection retains alternatives when multi-objective or uncertainty conditions do not justify one scalar winner.
- Quality-Diversity and archive-based evaluation retain the descriptor, distance and archive-policy editions their evaluation uses, so a telemetry change can be interpreted against that basis. An archive kept solely to preserve evidence needs no QD pins.
- Open-endedness / curriculum generation use explicit transfer rules and applicable degrade branches, with telemetry available to refresh orchestration.
- Reproducible publication keeps identifiers, editions, policies and traces recoverable for a later application or audit.
G.8:13 - Relations
Builds on: G.Core, C.23, G.4, G.6, G.5, C.22
Uses: A.10 (anchors), F.8 (policy-id resolvability), F.9 (cross-semantic relation and bounded-use claims), F.17/E.17 (crossing publication surfaces), G.7 (when calibration is used), B.3 (when a separate named assurance use is made), G.11 (refresh planning/trigger consumption), G.10 (shipping boundary; if bundled artefacts are shipped), E.10 (LEX twin registers), E.5.2 (notation independence), E.18 (crossing visibility when a selected transformation-flow structure is in use), A.21 (gate decisions under an applicable profile); optional C.18 (QD) / C.19 (E/E‑LOG) when those surfaces are declared.
Publishes to: UTS (bundle/ledger/card), G.5 (selector/registry consumption), G.11 (refresh via typed triggers and pinned telemetry)
Constrains: any SoS‑LOG packaging that claims FPF conformance for selector‑facing dispatch across method families.
G.8:14 - Author’s quick checklist (informative)
-
List the IDs of the applicable C.23 rules in
SoSLogRuleId[]. -
Any numeric gates/thresholds are
G.4Acceptance artefacts cited by id (no thresholds embedded in LOG or rungs). -
Evidence is citable: at run time use
PathId/PathSliceIdwhen available; at packaging time provide resolvableA10EvidenceGraphRef?[]/EvidenceGraphId?. -
Every actual sense, kind or plane correspondence retains its relation and receiving-use grounds under G.Core:4.2.3. For F.9 reuse,
G.8:Ext.BridgeReuseWiringkeeps the Bridge, separate bounded-use claim and reliance basis citable.CL/CL^k/CL^planeandΦ/Ψ/Φ_planeids and editions are pinned when the calibration or named assurance account requires them; policy ids resolve underF.8:8.1. -
PortfolioModeand dominance defaults are not restated: cite each default’s governing definition throughG.Core.DefaultGoverningDefinitionIndex(governing definitions live outsideG.8, typicallyG.5). -
When the bundle binds a QD use, its required descriptor, distance, insertion and emitter definitions are edition/policy pinned (
DescriptorMapRef.edition,DistanceDefRef.edition, insertion/emitter policies); the exact space and its separate description-episteme reference are pinned iff cell boundaries/de-dup/parity depend on the space definition; Spaces ≠ Maps.PortfolioMode=Archivealone does not activate these QD fields. -
If open-ended surfaces are declared, pin
GeneratorFamilyRowRef,TransferRulesRef.editionand any validity/coupler policies. Handle unknown transfer under the declared C.23 branch: permitteddegradeafter its prerequisites pass, otherwiseabstain; report unavailable basis and unperformed judgements under C.23:4.2.1. -
MaturityRungsis a closed, UTS‑registered set; the maturity ladder is ordinal/poset with a declaredReferencePlane; rung transitions cite evidence. -
RSCR triggers are emitted as canonical
RSCRTriggerKindIdvalues (no prose-only “reasons”). -
Notation independence (
E.5.2) and twin‑register discipline (E.10) are respected for all published heads/ids. -
If authoring tools materially shaped rule/rung content, cite
AuthoringMethodDescriptionRefs?[](edition‑pinned) for cross‑stance traceability.
G.8:End
G.9 — Parity and Benchmark Harness
Status: Stable
G.9:0 — Use this when
- rival method families, method sets, or adaptation paths must be compared under one declared baseline set and freshness window
- you need parity to publish one reproducible report rather than one opaque benchmark score
- downstream selection must recover comparator, normalization, bridge, and evidence pins without relying on one hidden scoring sheet
First move. State whether the comparison concerns exact targets or results about one subject, then fix the baseline, comparison rule and windows in one plan edition. Use A.19.CPM directly for a bounded comparison that needs no reproducible parity report; use this harness when the report and its replay basis are required.
G.9:0.1 — What goes wrong if missed
- benchmark reports present numbers from different windows, baselines, or comparator editions as comparable
- reuse across distinct source-local meanings, a reference-plane crossing, or a normalization mapping stays hidden until a disagreement appears downstream
- parity flattens a partial order into one scalar winner and silently changes what the comparison means
G.9:0.2 — What this buys
- one
ParityPlanRefthat fixes the plan edition, baseline, freshness, comparator, and bridge discipline up front - one
ParityReportthat cites that plan edition and echoes its active baseline binding, pins, outcomes, and evidence trace by value - one harness that downstream selection can consume without inventing a
G.9-local CSLC gate or a shadow governance card
Illumination, coverage and regret are telemetry by default. To use them in dominance, declare the policy that authorizes the promotion and record its ID.
G.9:1 — Intent
Provide a notation‑independent harness that:
- plans parity runs for one explicit subject—either one
EntityOfConcernRefor target refs under their existing subject patterns—with aReferencePlane, scope, window, applicable rules, CSLC comparability and admissibility references, comparator references (CNSpecRef,CGSpecRef,ComparatorSpecRef), and reproducibility pins for editions and policy ids; - executes the comparison with a DRR and SCR evidence trace; when selection is current, G.5 consumes its result and declares the selector outcome;
- publishes an edition-pinned ParityReport suitable for downstream consumption, shipping, refresh wiring, and RSCR.
G.9:2 — Problem frame
Parity claims become non‑reproducible or non‑comparable when any of the following are implicit:
- evidence window and freshness regime,
- comparator semantics, including any normalization or comparability mapping,
- the active C.21 replay basis when DHC coordinates are compared, including the exact Characteristic, Scale, measurement definition, Method, MethodDescription or model, and time or population basis,
- reuse across distinct F.17 source-local meanings or ReferencePlanes (the obtaining relation, crossing pins, and CL penalty placement),
- dominance and
PortfolioModeinterpretation rules, - gate outcomes (why a run abstained or degraded).
G.9 publishes the references needed to recover these inputs and outcomes when disclosing how the comparison was obtained (MOO). This disclosure cites the governing specifications without introducing a new governing specification reference.
G.9:3 — Forces
- Pluralism vs comparability. Support justified comparisons across Traditions while preserving their differences in meaning and admissibility.
- Partial orders. Many targets are only partially ordered; parity reporting must preserve CSLC-admissible outcome shape (often declared set results rather than a single scalar; an archive may supply the compared candidates).
- Edition sensitivity. A reader must be able to tell which measurement and comparator definitions a parity run used, even after those definitions change. When DHC, QD, or OEE modes are used, the required definition pins are introduced only through the corresponding
Extensionsblocks; omit them when unused. - Telemetry versus objectives.
IlluminationSummary, coverage, and regret are report-only telemetry by default. A dominance change needs an explicit CAL policy id recorded in the audit pins. - Crossing visibility. An actual E.18 flow crossing or A.21 gate used by parity retains its required
CrossingBundle, pins and checks. Failure blocks the dependent publication or use. - Cross-sense and reference-plane reuse. When expressions have distinct F.17 source-local meanings, recover both cells and establish the required F.9 relation; a ReferencePlane crossing follows its own declared crossing basis. Each actual correspondence retains its required pins and evidence; any justified numerical loss follows its receiving model and R-channel placement.
- Refreshability. Parity must emit RSCR-relevant causes as canonical ids, with enough pins to identify the affected use and plan any needed rerun.
G.9:4 — Solution
G.9:4.0 — G.Core linkage (normative)
This pattern binds to G.Core through the following manifest.
GCoreLinkageManifest (G.9) (normative; expands per G.Core:4.2)
Effective obligations/pins/triggers are computed as union(expand(sets), explicit deltas) under Nil‑elision.
-
CoreConformanceProfileIds:= {GCoreConformanceProfileId.PartG.AuthoringBase,GCoreConformanceProfileId.PartG.TriStateGuard,GCoreConformanceProfileId.PartG.ShippingBoundary,GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted} -
RSCRTriggerSetIds:= {GCoreTriggerSetId.CGSpecGate} -
RSCRTriggerKindIds:= {RSCRTriggerKindId.EvidenceSurfaceEdit,RSCRTriggerKindId.PenaltyPolicyEdit,RSCRTriggerKindId.BaselineBindingEdit,RSCRTriggerKindId.TelemetryDelta} (Pattern-local deltas; cross-tradition or Bridge-calibration causes are wired viaG.9:Ext.CrossTraditionParityand MUST NOT over-trigger parity runs that use one already recovered meaning and ReferencePlane.) -
DefaultsConsumed:= {DefaultId.DominanceRegime,DefaultId.PortfolioMode,DefaultId.GammaFoldForR_eff} (Use each default when its operation needs it. G.Core.DefaultGoverningDefinitionIndex identifies the governing definitions CC‑G5.28, CC‑G5.23 and CC‑G5.4 respectively.) -
CorePinSetIds:= {GCorePinSetId.PartG.AuthoringMinimal,GCorePinSetId.PartG.CrossingVisibilityPins} -
CorePinsRequired(pattern delta; pin names only; all are id‑valued unless noted) := {ComparatorSpecRef.edition,TaskSignatureRef?,TaskMapRef?, (when a selector consumes the parity result; TaskMapRef only when its G.4 CAL gates are current)entityOfConcernRef?,targetRefs[]?, (exactly one subject branch)ClaimScope,EvaluationWindow,FreshnessWindows,BaselineSet,BaselineBindingRef,ParityPinSet,PlannedFillingRows[]?,EvidenceGraphId,Budgeting?,EpsilonDominance?,UNM_id?,NormalizationMethodId[]?,NormalizationMethodInstanceId[]?,SCPRef.edition?,MinimalEvidenceRef.edition?} (Nil‑elision applies; mode‑specific definition pins are introduced only by the correspondingGPatternExtensionblocks.) -
TriggerAliasMapRef:=∅
G.9:4.1 — Objects and publication records
All objects below are notation‑independent; serialisations (if any) are handled in shipping and interop publication forms, not here.
(1) ParityPlan (one U.WorkPlan episteme; ParityPlan is the local application name)
A plan that fixes what is being compared and under what pinned conditions.
Minimal fields (conceptual; ids/pins only):
ParityPlan := ⟨
ParityPlanId(UTS), // continuing plan lineage
planEdition, // one immutable edition
CGFrameId?, // exact cited CG frame when the plan depends on one
TaskSignatureRef?, TaskMapRef?, // conditional G.5 input; TaskMap only for current G.4 CAL gates
entityOfConcernRef? := EntityOfConcernRef, // one-EntityOfConcern branch only
targetRefs[]?, // exact-target branch only; existing kinds and editions
groundingHolonRef?, // required for the empirical grounding used by this claim
referencePlaneRef := ReferencePlane,
claimScopeRef := ClaimScope,
EvaluationWindow,
UNM_id?, NormalizationMethodId[]?, NormalizationMethodInstanceId[]?, // when “normalize, then compare” is required (ids only; semantics come from CN‑Spec / UNM)
EpsilonDominance?, // optional ε-front thinning (ε≥0; id/param; pinned when used)
PortfolioMode?, DominanceRegime?, // cite each when the operation uses it, explicitly or through its governing default
BaselineSet, // exact method-family or generator-family targets (ids; notation-independent)
BaselineBindingRef, // evidence-backed baseline-set reference that says what counts as baseline
FreshnessWindows,
CNSpecRef.edition, CGSpecRef.edition, ComparatorSpecRef.edition, // edition-pinned refs
SCPRef.edition?, // optional (when a specific SCP profile must be pinned/cited)
MinimalEvidenceRef.edition?, // optional (when CG-Spec exposes minima profiles by ref)
Budgeting?,
ParityPinSet,
EvidenceGraphId, PathId[], PathSliceId?,
PlannedFillingRows[]? // declaration-local A.15.3 content inside this WorkPlan; no independent row refs
⟩
ParityPlanRef := <ParityPlanId, planEdition> designates one immutable plan edition. Changing its subject, baseline binding, comparator edition, or another active value that can change the run or its interpretation creates a new planEdition.
When a G.5 selector consumes this comparison, pin its TaskSignatureRef in the plan. If that selector uses G.4 CAL gates, also pin its TaskMapRef and require the map’s task-signature reference to match. Omit these selector inputs from a parity run that needs no selector.
A later edition may retain the lineage ID only if it continues the same plan. Earlier ParityPlanRef values still resolve their original editions.
Exactly one subject branch is present. Use entityOfConcernRef when the report compares results about one EntityOfConcern. Use targetRefs[] when the targets themselves are compared; each ref keeps the kind and edition defined by its existing subject pattern. In particular, a G.5 method-family target is an exact MethodFamilyRowRef, and a generator-family target is an exact GeneratorFamilyRowRef.
For example, a comparison of <ThresholdTrendReview-local, R3> and <SpectralResidualReview-local, R2> puts those two row refs in targetRefs[]; the plan may explicitly use the same two refs as its BaselineSet. A comparison of their results for Pump-P17 instead puts Pump-P17 in entityOfConcernRef, leaves targetRefs[] absent, and uses BaselineBindingRef to say how the two method rows supply results about that pump.
BaselineSet names the alternatives treated as the comparison baseline; it supplies targetRefs[] only when the plan explicitly says that the same exact refs serve both purposes. Otherwise the subject and baseline remain separate, and BaselineBindingRef records how that baseline applies to the named subject.
These values determine what is compared and when the parity claim is usable; do not add ParityContextId. If the plan relates expressions with distinct source-local meanings, first recover the exact F.17 cells and establish the required F.9 relation. A shared label, source note, or generic context identifier does not establish comparability.
(2) ParityPinSet (pin set)
A declared set of pins required for reproducibility and audit (editions + policy‑ids + UTS/Path pins).
G.9 specifies the contents of ParityPinSet under the common pin rules in G.Core.
(3) ParityReport (UTS publication record; work-result or audit-facing publication record only when the neighboring source exists)
A UTS-publishable parity publication record produced by a parity run under one ParityPlanRef. Work or audit occurrence claims use A.15/A.15.1; evidence-path claims use A.10/G.6; a separate named assurance claim uses B.3; and a gate decision uses A.21 under its applicable profile. Keep each such occurrence, path, or result distinct from the report.
ParityReport := ⟨
ParityReportId(UTS),
parityPlanRef := ParityPlanRef,
entityOfConcernRef?, targetRefs[]?, // exactly one subject branch is present
TaskSignatureRef?, TaskMapRef?, // echoed when the selector branch is used
groundingHolonRef?, referencePlaneRef, // echo the empirical ground when used
claimScopeRef, EvaluationWindow,
BaselineSet, BaselineBindingRef, FreshnessWindows,
CNSpecRef.edition, CGSpecRef.edition, ComparatorSpecRef.edition,
SCPRef.edition?, MinimalEvidenceRef.edition?, // echoed iff used/pinned in the plan
UNM_id?, NormalizationMethodId[]?, NormalizationMethodInstanceId[]?, // echoed iff used in the plan
OutcomeRefs, // parity/comparison results; G.5 output refs only when selection is current
EpsilonDominance?, // echoed when used
AbstainReasons[]?, // ids/labels (policy-bound) for abstain/degrade; refusal paths included
TelemetrySummary? := ⟨IlluminationSummary?, coverage?, regret?⟩, // report-only by default; promotion requires CAL policy-id pins
GuardOutcomeTraceRef?, // pass/degrade/abstain trace + cited reasons (policy-bound)
EvidenceTrace := ⟨EvidenceGraphId, PathId[], PathSliceId?⟩,
CrossingPins?, // Bridge/CL/Φ/Ψ/Φ_plane pins, when crossings are invoked
EditionPinsDelta?, // explicit list of edition pins actually active during the run
PolicyPinsDelta?, // explicit list of policy-ids actually active during the run
RSCRRefs[] // parity RSCR test ids / trigger emissions
⟩
The report carries the ParityPlanRef and echoes the BaselineBindingRef used in that edition. For example, if <PumpParityPlan, E4> used PumpBaselineBinding-E7 and a later E5 changes the binding or comparator, an old report still resolves E4 and PumpBaselineBinding-E7. A missing historical plan edition or binding is an unresolved required input; it is never replaced with the current value.
Naming discipline.
- Head names follow the existing kind definitions and LEX discipline.
- The older labels
ParityPlan@ContextandParityReport@Contextare retired. The suffix did not identify the object or the comparison basis. UseParityPlanandParityReport, with all operative conditions carried in explicit fields and exact refs. - Tech/Plain twins follow E.10 rules (no drift‑inducing synonyms in Tech).
G.9:4.2 — Parity planning (one U.WorkPlan)
Planning records the conditions needed to repeat and interpret the parity run:
-
Fix the baseline set. Choose the exact
BaselineSet(MethodFamilies, and optionally GeneratorFamilies) used as the comparison baseline. When SoS-log or source-maturity values change baseline eligibility or interpretation, citeSoS‑LOGBundleId?and the source-maturity ids by reference; acceptance-gate thresholds remain inG.4Acceptance. -
Bind subject, scope, and evaluation window. Choose exactly one subject branch: one
entityOfConcernRef, or exacttargetRefs[]under their existing kinds and editions. For G.5 families, useMethodFamilyRowReforGeneratorFamilyRowRef, not a bare lineage id. Record theReferencePlane,ClaimScopeandEvaluationWindow. An empirical claim identifies itsgroundingHolonRef. For a comparison of two proven operation-count bounds on a formal domain, identify the bounds, their proofs and the domain. A computer that stores a proof supplies no empirical premise merely by storing it. -
Define baseline-set reference. Declare what counts as the baseline and how it applies to the selected subject in
BaselineBindingRef(for example, through an EvidenceGraph path slice or an upstream shipped package or publication-record id). IfBaselineSetalso supplies the exact compared targets, say so and use the same refs by value; otherwise keep baseline and subject refs distinct. -
Equalise window (and budget, if pinned). Declare a single
FreshnessWindowsand apply it across all baselines; ifBudgetingis used/pinned, it MUST be shared/pinned across baselines as well.When specialization is part of the parity claim, the same plan should also hold constant the declared task family or target scope cut, the work-measure threshold target, adaptation budget, prior exposure declaration, and freshness window. Identify what each candidate performed and what assistance it had. To compare the candidates’ own acquisition, account for differences in support that could explain the result. If the arrangements themselves are being compared, retain each arrangement as the bearer of its performance. When transfer, retention, later efficiency, downside or corridor entry matters, keep the observations and the baseline supporting that claim under C.22.1:8.
-
Pin governance, CSLC comparability and admissibility references, and comparator references.
CNSpecRef,CGSpecRef, andComparatorSpecRefare referenced with explicit edition pins. -
Pin measurement/comparator definitions (conditional). Where parity depends on mode‑specific definition records (e.g., DHC/QD/OEE), pin the relevant definition ids/editions/policies. The minimum required pins are declared by the applicable
Extensionsblocks (e.g.,G.9:Ext.DHCParityPins,G.9:Ext.QDArchiveParity,G.9:Ext.OEEParity) and the referenced records they cite. -
Bind comparator choice to CG-Spec. Any numeric comparison or aggregation MUST be CSLC-admissible and cite the corresponding CG-Spec entry (
ComparatorSpecRef). Compare corresponding coordinates on compatible scales and bases. A Pareto comparator can compare kWh with kWh and minutes with minutes in their own units. When a coordinate needs conversion or the admitted operation uses normalization, declare the applicableUNM_id?,NormalizationMethodId[]?andNormalizationMethodInstanceId[]?and establish that the mapping preserves the distinctions the comparator uses. -
Declare the required result. An admitted A.19.CPM comparison returns relation or poset tokens. Preserve established incomparability.
CompareEligibilitycan also withhold comparison if required evidence is missing. Pin the dominance regime when using dominance, including for a frontier. Pin the portfolio mode when the operation uses one. Each can cite its applicable default throughG.Core.DefaultGoverningDefinitionIndex. IlluminationSummary, coverage and regret remain telemetry unless a cited CAL policy promotes them. -
Attach planned fillings when applicable. If parity depends on planned slot fillings, this WorkPlan contains the relevant A.15.3 rows in
PlannedFillingRows[]; each row points to a declaration member defined by its own pattern and has no independent reference or identity. Omit the field when no such row is needed. -
Publish the relation and its receiving use (when invoked). For distinct recovered F.17 meanings, establish the F.9 relation and the separate bounded-use claim and matching reliance. Cite G.7 calibration, CL and policy pins when that account uses them; its summary does not decide parity admissibility. Kind correspondence also requires receiving admissibility and fresh target classification under C.3.3. A plane relation keeps its own basis. Any numerical assurance penalty follows the applicable B.3 model for the named use.
G.9:4.3 — Execution protocol (run‑time / selector‑adjacent)
For one run under the pinned plan, follow these steps:
-
Validate CSLC references and pins. Validate the cited CSLC comparability and admissibility references, active pins, and witnesses. Apply the eligibility or acceptance checks required by the pinned comparison and refuse or abstain on non-admissible operations. When using G.5, record its branch trace in its own result vocabulary and use the planned
TaskSignatureRef. Add the matchingTaskMapRefwhen that selector uses G.4 CAL gates. If a liveA.21gate consumes this check, cite itsGateDecisionResultand add aDecisionLogonly when needed for audit or reuse. The profile application and its action consequence remain governed by A.21; do not create aG.9-local CSLC gate. -
Run the comparison and any required selection. Apply the pinned comparator to the admissible comparison inputs and retain its lawful result shape in
ParityReport. If a selector-facing result is required, apply G.5 under the plan’s exact task and conditional map refs and publish its declared outcome. An ordinary parity or benchmark report ends without G.5 when no selector result is needed.Under componentwise minimization, A=(8,3) and B=(10,2) trade cost against time. An admitted comparison can return the singleton
{A ∥ B}. If A and B are all eligible candidates, both are nondominated. Add C=(7,2) and the frontier becomes{C}while the relation between A and B stays the same. Under a comparator that admits intervals, take A=(8,[3,4]) and B=(10,2). Every allowed value for A’s time still gives incomparability. With A’s time in [1,4], A may dominate B or the two may trade off. Report only what the approved comparison supports. If two bounds used as evidence about the same profile contradict each other, resolve that conflict before reporting a relation. If the profiles are consistent and none meets the requested limits, that can settle the eligibility question. Separately, if a required calibration is missing, apply the evidence policy. An admitted narrower comparison can return only its justified tokens. Anabstaindecision leaves the comparison slot unbound.For a bounded-specialization comparison, the report should include the active specialization profiles or equivalent pins so readers can recover the work-measure threshold target, prior exposure, budget-to-threshold, post-threshold efficiency when relevant, transfer, retention, downside field, and any corridor-entry baseline or evidence note from the report.
-
Record the comparability mapping when used. If
UNM_id?,NormalizationMethodId[]?, orNormalizationMethodInstanceId[]?was declared, echo it inParityReportor its explicit pins delta. Record the ids and any scoped notes required by the cited specification in the audit pins and SCR; cite the applicablePathIdvalues. -
Publish trace. Emit
ParityReportwith theParityPlanRef, itsBaselineBindingRef, conditional selector input refs, EvidenceGraph citations, and all active edition and policy-id pins, so the run can be checked and run again. -
Emit telemetry for refresh when produced. Publish the telemetry pins or events used for refresh. Treat the values as report-only unless a CAL policy promotes them and its ID is recorded.
G.9:4.3a — Worked parity slice
Ordinary case: compare two pump-triage method rows. The team from the G.5 example wants a reproducible comparison of the two exact selector-row editions, not a claim that one Method is universally better. Both rows use the same scheme, units, and ReferencePlane, so no normalization or crossing branch is needed.
ParityPlanRef = <PumpTriageParity, E1>
targetRefs = [<ThresholdTrendReview-local, R3>,
<SpectralResidualReview-local, R2>]
entityOfConcernRef = absent
groundingHolonRef = PumpMaintenanceProgram-H1
referencePlaneRef = PumpVibrationTriage-RP1
claimScopeRef = PumpFleet-F7-VibrationTriageClaims-E1
EvaluationWindow = 2026-08-01T00:00Z .. 2026-08-07T23:59Z
BaselineSet = [<ThresholdTrendReview-local, R3>,
<SpectralResidualReview-local, R2>]
BaselineBindingRef = PumpTriageBaselineBinding-E1
FreshnessWindows = { sensorSeries: at-most-24h-old-at-run,
evidencePath: at-most-72h-old-at-run }
CNSpecRef.edition = PumpCN-E2
CGSpecRef.edition = PumpCG-E4
ComparatorSpecRef.edition = PumpTriageComparator-E3
TaskSignatureRef = PumpVibrationTriage-T1
TaskMapRef = absent // this G.5 case uses no G.4 CAL gate
ParityPinSet = [PumpCN-E2, PumpCG-E4, PumpTriageComparator-E3, PumpVibrationMeasureSpec-E2]
EvidenceGraphId = PumpTriageEvidence-E5
PathId[] = [PumpReadings-P7, ComparatorRun-P3]
PathSliceId = PumpParitySlice-S2
expected selector result = unordered Shortlist
The plan explicitly says that BaselineSet and targetRefs[] contain the same two row refs. EvaluationWindow bounds the observations and results included in the comparison. FreshnessWindows asks a different question at run or reuse time: whether each required input and evidence path is still recent enough to rely on.
A report from this run carries parityPlanRef=<PumpTriageParity, E1>, BaselineBindingRef=PumpTriageBaselineBinding-E1, the same evidence path, and the unordered Shortlist; it does not invent a scalar winner.
Conditional specialization case. Loop-engineering parity may add further pins after the ordinary comparison boundary above is complete. An evaluation program, benchmark script, or dashboard is part of the evaluation or comparison procedure; it is not the Characteristic being improved.
- Two agentic search setups both claim bounded specialization on the same declared task family.
- Their
ParityPlanalso pins the same threshold target, adaptation budget, prior-exposure declaration, and corridor-entry baseline. One setup reaches threshold sooner but shows low retention and no transfer. The other reaches threshold later, but carries reusable transfer and lower downside field. - Their CSLC-admissible
ParityReportstates what was held constant, which signals remained telemetry, and why the outcome stays a selected set or partial order rather than collapsing into a scalar winner. These specialization values extend the complete comparison boundary; they do not replace its subject, scope, windows, baseline binding, comparator editions, or evidence.
G.9:4.3b — Conditional causal method rung parity
Use this extension only when a parity report compares causal methods or causal-use claims. Start with a cheap screen and stop at degraded parity or abstention when the methods answer different questions.
CausalRungParityScreen:
comparedMethodsRef
targetCausalityLadderRungSet
causalSupportComponentTypeSet
sameEstimand: yes | no | unclear
sameOutcomeWindow: yes | no | unclear
sameTransportEndpoints: yes | no | unclear
cheapParityStop:
comparableEnoughForFullRecord |
crossRungDegrade |
crossSupportComponentsDegrade |
differentEstimandAbstain |
differentOutcomeWindowAbstain |
differentEndpointsAbstain |
returnToC28
causalSupportComponentTypeSet records which methods rely on evidence paths/data regimes, identification, estimates, direct counterfactual sampling, simulation, or transport. Expose the difference and justify the proposed comparison under its causal-support conditions; a shared label does not make unlike components equivalent.
Open the full record only when comparison remains meaningful:
CausalMethodRungParityRecord:
comparedMethodsRef
causalUseQuestionRef?: CausalUseQuestionRef
targetCausalUseClaimKind: CausalUseClaimKind
targetCausalityLadderRung: CausalityLadderRung
causalEstimandRef: CausalEstimandRef
declaredCausalityLadderBridgeOrLossRef?
interventionBudgetOrActionSetRef?
causalSupportComponentRefs: CausalSupportComponentRefs
declaredCausalSupportLossRef?
causalUseSupportResultRef?: CausalUseSupportResultRef
causalFollowUpWindowRef
outcomeMeasureRef
sourcePopulationRef?
targetPopulationRef?
sourceDomainRef?
targetDomainRef?
sourceEnvironmentRef?
targetEnvironmentRef?
sourceDataGeneratingRegimeRef?
targetDataGeneratingRegimeRef?
transportabilityResultRef?
estimateResultRef?
parityVerdict: parityEstablished | degraded | abstain
supportedParityUse
unsupportedParityUse
The record names every changed transport endpoint that matters; population and semantic scheme do not substitute for domain, environment, or data-generating regime. Different rungs, estimands, windows, endpoints, or support components require a bridge/loss, degraded parity, or abstention. G.9 makes the parity conclusion; C.28 supplies the cited causal-support result and does not authorize the benchmark conclusion.
G.9:4.9 — Extensions (pattern‑scoped; non‑core)
For an ordinary parity comparison that uses none of these modes, you can stop after the worked case in §4.3a. Use §4.3b when comparing causal methods or causal-use claims, and the extension blocks below when their mode is active.
The extensions specify only the pins, references, policy IDs, triggers and links needed for each mode. The referenced patterns define what these mean.
GPatternExtension block: G.9:Ext.CrossTraditionParity
GPatternExtension: CrossTraditionParity
- PatternScopeId:
G.9:Ext.CrossTraditionParity - GPatternExtensionId:
CrossTraditionParity - GPatternExtensionKind:
DisciplineSpecific - GoverningPatternId:
G.7 - Uses:
{G.7, F.9, E.18, A.21} - ⊑/⊑⁺:
∅ - RequiredPins/EditionPins/PolicyPins (minimum; conditional on use):
BridgeId/BridgeCardId[](for an actual F.9 sense correspondence; cite a card when used)BridgeMatrixId?CalibrationLedgerId?/BCT.id?RegressionSetId?/SentinelId[]?(when sentinel wiring is used)CL/CL^k/CL^plane(only the channel values consumed or required by the calibration or named assurance account)Φ(CL) policy-id,Φ_plane policy-id,Ψ(CL^k) policy-id?(only when the receiving model or assurance account requires that policy)CrossingBundleId?(required for an actual flow/gate use whose governing rule requires the bundle)
- RSCRTriggerSetIds:
{GCoreTriggerSetId.BridgeCalibrationKit}(preferred; expands inG.Core) - RSCRTriggerKindIds (delta, if any):
∅ - Notes: Apply the channel and receiving-use conditions of G.Core:4.2.3. A kind or plane correspondence keeps its own relation and receiving grounds; it creates no F.9 Bridge. The pins cite the calibration, loss model or crossing bundle consumed.
GPatternExtension block: G.9:Ext.SoSLogGuardNarration
GPatternExtension: SoSLogGuardNarration
- PatternScopeId:
G.9:Ext.SoSLogGuardNarration - GPatternExtensionId:
SoSLogGuardNarration - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
C.23 - Uses:
{C.23, G.6, G.4} - ⊑/⊑⁺:
∅ - RequiredPins/EditionPins/PolicyPins (minimum; conditional on use):
SoSLogRuleId[]/BranchId[](ids as cited labels; semantics come fromC.23)FailureBehaviorPolicyId/SoSLogBranchIdEvidenceTrace.PathId[]/PathSliceId?AcceptanceClauseId[](when referenced)
- RSCRTriggerKindIds:
{RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.MaturityRungChange, RSCRTriggerKindId.TelemetryDelta} - Notes (wiring-only): Explains why a parity run degraded/abstained by citing SoS‑LOG ids and evidence paths; does not redefine guard semantics.
GPatternExtension block: G.9:Ext.DHCParityPins
GPatternExtension: DHCParityPins
- PatternScopeId:
G.9:Ext.DHCParityPins - GPatternExtensionId:
DHCParityPins - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
C.21 - Uses:
{C.21} - ⊑/⊑⁺:
∅ - Required replay values and pins (minimum; conditional on DHC parity):
DisciplineRefIntendedUseClaimScopeRefComparisonBasisCharacteristicRef.editionScaleRef.editionUnitRef.edition?DHCMethodRef.editionMethodRefMethodDescriptionRef.edition?MeasurementModelRef.edition?CalibrationBasisRef?TimeOrPopulationBasisDHCDefinitionSetRef.edition?TargetSliceRef?DistanceDefRef.edition?
- RSCRTriggerKindIds:
{RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.EvidenceSurfaceEdit} - Notes (wiring-only): Carry exactly the active fields of the C.21 replay basis.
TargetSliceRefappears only when the parity computation consumes that A.2.6 selection and states its relation toClaimScopeRef. Compatible same-semantics readings use the admitted C.16 comparison basis directly; actual distinct-local-sense use also cites the obtaining F.9 relation, direction, admitted use, and loss. C.21 defines the DHC semantics.
GPatternExtension block: G.9:Ext.QDArchiveParity
GPatternExtension: QDArchiveParity
- PatternScopeId:
G.9:Ext.QDArchiveParity - GPatternExtensionId:
QDArchiveParity - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
C.18 - Uses:
{C.18, C.19, G.5} - ⊑/⊑⁺:
∅ - RequiredPins/EditionPins/PolicyPins (minimum; conditional on use):
DescriptorMapRef.editionDistanceDefRef.editionCharacteristicSpaceRefwith separatecharacteristicSpaceDescriptionRef: U.EpistemeRef(when discretisation/topology is referenced; resolve the same-space pair under A.19:5.1.1.1)EmitterPolicyRefInsertionPolicyRef
- RSCRTriggerKindIds:
{RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta} - Notes: Pin the QD definitions used by the comparison; their semantics remain in
C.18,C.19andG.5.
GPatternExtension block: G.9:Ext.OEEParity
GPatternExtension: OEEParity
- PatternScopeId:
G.9:Ext.OEEParity - GPatternExtensionId:
OEEParity - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
C.19 - Uses:
{C.19, G.5} - ⊑/⊑⁺:
∅ - RequiredPins/EditionPins/PolicyPins (minimum; conditional on use):
TransferRulesRef.editionEnvironmentValidityRegionIdExplorationBudgetPolicyId?EvidenceTrace.PathSliceId?(for transfer‑keyed events)
- RSCRTriggerKindIds:
{RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta} - Notes (wiring-only): For open-ended parity, cite the policies and editions above and connect their telemetry to refresh. The extension adds no core rule.
G.9:5 — Interfaces (minimal I/O; conceptual)
| Interface | Consumes | Produces |
|---|---|---|
G.9‑1 Plan_Parity | exactly one subject branch—one EntityOfConcernRef or exact targetRefs[] under their existing kinds and editions—plus GroundingHolonRef when empirical grounding is used, ReferencePlane, ClaimScope, EvaluationWindow, BaselineSet, BaselineBindingRef, FreshnessWindows, Budgeting?, EpsilonDominance?, CNSpecRef.edition, CGSpecRef.edition, ComparatorSpecRef.edition, mode-specific measurement or normalization editions when used, SCPRef.edition?, MinimalEvidenceRef.edition?, UNM_id?, NormalizationMethodId[]?, NormalizationMethodInstanceId[]?, ParityPinSet, EvidenceGraphId, PathId[], PathSliceId?, PlannedFillingRows[]? | one immutable ParityPlan WorkPlan edition and its ParityPlanRef |
G.9‑2 Run_Parity | ParityPlanRef and comparison inputs; G.5‑3 Select only for a selector-facing result, using the plan’s TaskSignatureRef and conditional matching TaskMapRef | parity/comparison result refs and, when selected, G.5 outcome refs; DRR and SCR pins with PathId[] and, when needed, PathSliceId |
G.9‑3 Publish_ParityReport | ParityPlanRef, parity-run trace refs, and active pins | ParityReport carrying the same plan ref and baseline binding (UTS publication record; emits canonical RSCR ids) |
G.9‑4 Expose_ParityTelemetry | Telemetry deltas (archive changes, coverage/regret signals, etc.) | Telemetry events carrying PathSliceId?, policy‑ids, and edition pins for refresh wiring |
Publication records are conceptual here; serialisations belong in shipping and interop publication forms (see G.10 and interop annexes), not in G.9.
G.9:6 — Conformance Checklist (CC‑G9)
CC‑G9‑CoreRef (normative).
G.9 conforms only if it satisfies the effective set of CC‑GCORE‑* declared in G.9:4.0 GCoreLinkageManifest (including trigger typing, Default Governing Definition Index links, and P2W split).
-
CC‑G9.1 — Exact comparison boundary, equal windows (and budgets), and pinned spec editions (local). A
ParityPlanRef = <ParityPlanId, planEdition>SHALL resolve one immutable plan edition. That ParityPlan SHALL choose exactly one subject branch: oneEntityOfConcernRef, ortargetRefs[]under their existing kinds and editions. It SHALL nameGroundingHolonRefwhen empirical grounding is used. It SHALL also nameReferencePlane,ClaimScope,EvaluationWindow, baseline set and binding, and evidence refs, and SHALL declare a singleFreshnessWindowsshared across baselines.BaselineSetsupplies the target refs only when the plan explicitly identifies the same refs in both places; otherwiseBaselineBindingRefrelates the separate baseline to the named subject. IfBudgetingis used and pinned, it SHALL be shared across baselines as well.ParityPinSetSHALL include the editions required by the referenced specification, comparator, and any measurement or normalization method in use (at minimumCNSpecRef.edition,CGSpecRef.edition,ComparatorSpecRef.edition). If the parity run depends on planned slot fillings, its exactParityPlanWorkPlan SHALL carry the relevant declaration-local A.15.3 rows inPlannedFillingRows[](nil-elision when not applicable). Each row resolves only inside that WorkPlan and has no independent reference, kind, or edition. -
CC‑G9.2 — Mode‑specific definition pins are declared via Extensions (local; conditional). When parity depends on mode‑specific definition records beyond the pinned governing spec refs (e.g., DHC/QD/OEE), the ParityPlan/Report SHALL include the corresponding
GPatternExtensionblocks and satisfy theirRequiredPins/EditionPins/PolicyPins(typically carried insideParityPinSet, and echoed via pins deltas in audit):- DHC parity →
G.9:Ext.DHCParityPins - QD archive parity →
G.9:Ext.QDArchiveParity - OEE parity →
G.9:Ext.OEEParity
- DHC parity →
-
CC‑G9.3 — CSLC-admissible orders and arithmetic (delegation point + local constraint). Delegated to
CC‑GCORE‑SET‑1(and the relevant G.5PortfolioMode/ selected-set semantics). Additionally: any numeric comparison or aggregation invoked by parity SHALL be CSLC-admissible and cite the corresponding CG‑Spec entry; non-admissible operations (e.g., ordinal means / mixed‑scale weighted sums) SHALL be refused or abstained with path‑cited trace (citation only; arithmetic admissibility comes fromCG‑Spec/MM‑CHR). -
CC‑G9.4 — Normalization discipline (local citation only). When the comparator or an operand requires a CSLC-admissible comparability mapping, the ParityPlan SHALL cite the mapping and its applicable ids (
UNM_id?,NormalizationMethodId[]?,NormalizationMethodInstanceId[]?) and compare the mapped values. A componentwise comparison may use the native units of different characteristics when operands match within each coordinate. If such mapping ids are used, the ParityReport SHALL echo the same ids (directly or via explicit pins deltas) so the run is reproducible and auditable without unrecorded information. The harness SHALL NOT define a local mapping. -
CC‑G9.5 — Dominance/PortfolioMode interpretation & telemetry separation (local).
ParityPlanandParityReportSHALL identify the dominance regime for operations using dominance and the portfolio mode for operations using a portfolio. Cite each applicable reference or its governing default inG.Core.DefaultGoverningDefinitionIndex. Bind and record a policy for any non-default promotion.IlluminationSummary, coverage, and regret SHALL be treated as telemetry (report-only by default); any promotion into dominance is an explicitly pinned CAL policy and MUST be recorded in the audit pins and SCR.5a. CC‑G9.5a — Adaptation parity disclosure (local; conditional). When the parity claim concerns bounded specialization, the ParityPlan and ParityReport SHALL pin the declared task family or target scope cut, the work-measure threshold target, adaptation budget, prior exposure, and the performer and assistance relevant to the claimed capability. Retain the basis for any transfer, retention, later exploitation efficiency, downside or corridor-entry claim that affects the comparison.
-
CC‑G9.6 — Epsilon‑front thinning (local; conditional). If ε‑front thinning is used,
EpsilonDominance (ε≥0)SHALL be explicit in the plan/report and pinned (param/id) such that the same ε is reproducible. -
CC‑G9.7 — Crossing visibility (delegation point). Delegated to
CC‑GCORE‑CROSS‑1andCC‑GCORE‑PEN‑1. This item remains as a stable delegation point for Bridge and reference-plane crossing visibility plus R-channel penalty placement discipline. -
CC‑G9.8 — Report replay and evidence trace completeness (local). A ParityReport SHALL carry the
ParityPlanRefandBaselineBindingRefused for the run, echoTaskSignatureRefand any requiredTaskMapRefwhen the selector branch was used, and include an EvidenceTrace withEvidenceGraphIdand the relevantPathId[](andPathSliceId?when needed), covering inclusions, refusals, abstentions, and degradations. If the historical plan edition or binding cannot be resolved, return that unresolved input instead of substituting a current edition. -
CC‑G9.9 — Telemetry hooks are emitted with pins (local). When parity emits telemetry for refresh, emitted telemetry SHALL carry the active edition pins and policy‑ids needed to re‑run parity (including the active subset of
ParityPinSetrelevant to the emitted event). In particular, telemetry items SHOULD citePathSliceIdwhen available, and SHALL include the policy id governing the telemetry interpretation. Mode‑specific definition pins SHALL be included as declared by the activeExtensionsblocks (e.g.,G.9:Ext.QDArchiveParity,G.9:Ext.OEEParity, includingEnvironmentValidityRegionIdwhen OEE parity is in scope). -
CC‑G9.10 — RSCR parity tests are published (local). Parity publication SHALL include RSCR parity tests (via
F.15harness refs) that cover negative/refusal paths relevant to this plan (missing pins, edition drift, missing bridge calibration refs, etc.). -
CC‑G9.11 — GateCrossing visibility (delegation point). Delegated to
CC‑GCORE‑CROSS‑1and the applicable GateCrossing/CrossingBundle harness checks (E.18,A.21,F.9, and relevant Part G bridge or crossing wiring). -
CC‑G9.12 — Tech‑register lexical discipline (local). Tech prose and heads SHALL follow E.10: do not introduce drift‑prone primitives (e.g., “metric” as a Tech primitive); reference the source pattern’s canonical terms and pinned refs.
-
CC‑G9.13 — MOO disclosure for parity (local).
Run_Parity/Publish_ParityReportSHALL record the ParityHarness identity (UTS ids) and the active pins required to interpret the outcome (editions + policy‑ids), so the reader can recover the harness and pins from the parity record itself. -
CC-G9-CLP-1 - Causal method rung parity. For a report comparing causal methods, the comparison SHALL begin with
CausalRungParityScreen. When full parity remains plausible, the report SHALL declare target causality-ladder rung, causal-use claim kind,causalEstimandRef, interventional-action basis, causal support-component refs, exact transport endpoints and transportability result when needed, estimate result when needed, bridge and loss where rungs differ, andcausalUseSupportResultRefwhen relevant C.28 support is consumed. Where parity cannot be established, the report SHALL state degraded parity or abstention.
G.9:7 — Anti‑patterns and remedies
- AP‑1 Hidden edition drift. Remedy: require edition pins in
ParityPinSet; treat changes as RSCR‑relevant via canonical trigger kinds. - AP‑2 Baseline set is informal prose. Remedy: require
BaselineBindingRefand EvidenceTrace pins. - AP‑3 Comparator semantics are “whatever the code did”. Remedy:
ComparatorSpecRef.edition(and any normalization/comparability refs) must be cited and pinned. - AP‑4 Cross-sense or reference-plane reuse without its obtaining relation and visible pins. Remedy: recover the exact F.17 cells and cite the obtaining F.9 relation when local meanings differ; cite the exact reference-plane crossing basis and visibility records when planes differ (delegated to G.Core).
- AP‑5 Parity report becomes a hidden scoring sheet. Remedy: preserve CSLC-admissible outcome shape and keep telemetry as telemetry unless explicitly policy‑promoted by the governing policy pattern.
- AP‑6 “Metric” as a primitive in Tech. Remedy: name the exact
CharacteristicRef,ScaleRef,UnitRefwhen applicable,DHCMethodRef,MethodRef, and the C.16 reading content or its result episteme; addDistanceDefRefonly when used. “Metric” may appear only in Plain with a pointer to those canonical objects. - AP‑7 Hidden DHC replay drift. Remedy: carry every active field of the C.21 replay basis and refuse parity reuse when a required field is unresolved or differs across the compared readings. Register refresh tests only for a named receiver that consumes those changes.
G.9:8 — Archetypal grounding
Show-A — Multi-tradition parity for decision systems. ParityPlan pins the evidence window and comparator references. ParityReport publishes what the comparison established and the supporting evidence. A commissioned selection adds its declared outcome. Preference-learning comparators, causal decision pipelines, offline-RL evaluation pipelines and robust BO-style selectors remain illustrative families until a G.2 SoTA pack or named current source identifies the family actually compared.
Show-B — QD parity (for example, CMA-MAE arXiv:2205.10752 and DQD arXiv:2106.03894).
ParityPlan pins descriptor/distance definitions and archive insertion policy editions. ParityReport identifies the compared archive states, records the lawful comparison result and telemetry deltas, and keeps illumination summaries separate from dominance unless explicitly promoted.
Show-C — Open-ended parity (POET arXiv:1901.01753 as an environment–method co-evolution lineage).
ParityPlan pins transfer rule editions and exploration policy refs. ParityReport publishes the comparison and any declared selected-set outcome with transfer-keyed traces (PathSlice). A policy change identifies the dependent uses for G.11 to reconsider.
Show-D — Causal method rung parity.
A team compares an observational predictor, an intervention optimizer, and a counterfactual policy strategy under one “best causal method” headline. The team first applies CausalRungParityScreen. If differences in rungs, support components, estimands, endpoints or outcome windows make the proposed comparison untenable, it reports degraded parity or abstention without opening the full record.
When full parity remains plausible, G.9 requires CausalMethodRungParityRecord. For each method, the record states targetCausalUseClaimKind, target CausalityLadderRung, causalEstimandRef, interventional-action basis, the support components consumed, relevant C.28 support result, follow-up window, outcome measure, changed transport endpoints, and estimate-result basis.
When fields differ, the report names the applicable declaredCausalityLadderBridgeOrLossRef and any available transportability or estimation references. It reports degraded parity or abstention where parity cannot be established. The admissible output may be a selected set by comparable rung, not one scalar winner.
G.9:9 — Cited Records (what this pattern publishes)
Exports (UTS‑publishable, edition‑pinned):
ParityPlanand itsParityPlanRef(oneU.WorkPlanepisteme and immutable edition reference; any planned-filling rows remain declaration-local content)ParityReport(UTS publication record carrying the plan and baseline-binding refs; work-result or audit-facing publication record only when the neighboring source relation is live)- DRR and SCR refs by id and, when applicable,
PortfolioPackRef?and selector-output refs by id, for downstream consumption. - Telemetry pins and events by id, for refresh wiring (
G.11) and RSCR harnesses (F.15).
G.9:10 — Relations
C.27 temporal-claim relation.
- C.27 may flag: dynamic parity when a benchmark compares rate-change, rhythm change, recovery speed, intervention effect, effort budget, or dynamic outcome.
- This pattern keeps: baseline, freshness, comparator edition, effort/budget parity, bridge discipline, parity plan, parity report, and reproducible benchmark publication.
- Non-admissible use: faster improvement is not benchmark superiority, and
dyn2BenchmarkParityBlock?is a benchmark input declaration, not a benchmark harness. - Exit: when live, recover
dynOrderCompared, baseline window, adaptation or intervention window, effort or budget parity reference, rate or rate-change measure,G9ParityPlanRef, and optionalG9ParityReportRef; G.5 is relevant only if a selector-facing result declaration consumes such a benchmark result.
C.29 mathematical-lens use relation.
- C.29 may flag: parity or benchmark input whose comparator, distance, descriptor geometry, embedding, normalization, surrogate model, learned representation, parity measure, model-family label, or model-selection basis depends on a mathematical lens that changes the parity claim and is missing, under-specified, or overread.
- This pattern keeps: baseline set, freshness, comparator edition, normalization ids, bridge discipline, parity plan, parity report, and reproducible benchmark publication.
- The
C.29mathematical-lens account is a cited input to parity.G.9governs the benchmark report and conclusion, andG.5governs any selector-facing output declaration. Parity-measure admissibility comes from the citedCG‑Spec/MM‑CHR. - C.29 application: for an under-lensed or overread parity input, cite the applicable
C.29output for the stated use:NoMathLensUseNeeded,MathLensUse.LensCandidateNote,MathLensUse.OneLine,MathLensUse.MiniCard,MathLensUse.FullCard, orNeighborGoverningPatternNote. Use the least costly adequate output that changes the next admissible parity use; full-card work is only required when the live parity or benchmark claim needs it.
Builds on: G.Core, G.5, G.6, G.4, F.15, E.17, E.18, A.21, F.17, E.5.2, E.10.
Publishes to: UTS (plan/report ids), G.11 (refresh wiring), G.10 (shipping publication form; parity records are cited records).
Uses: G.0, A.19, A.2.6 for exact U.ClaimScope, F.9, and C.28 when parity compares causal methods or causal-use claims.
Uses (optional, via Extensions): G.7, C.18 and C.19 (QD/OEE wiring), C.23 (SoS‑LOG narration and failure‑policy pins).
G.9:11 — Working reading checks
- If two baselines are being compared under different freshness windows, comparator editions, or silent normalization rules, this pattern has not yet been satisfied.
- If parity cannot tell the reader what was held constant, what remained telemetry, and what crossings or penalties were active, the report is not yet usable.
- If a scalar winner is being claimed where only a selected set or partial order is CSLC-admissible, parity is overclaiming and should publish the CSLC-admissible outcome shape instead.
G.9:End
G.10 - SoTA Pack Shipping
Status: Stable
Use this when. Part-G output must reach a selector, auditor or maintainer as one citable, editioned shipment. Start from that receiving use, collect the required upstream references and preserve the conditions for interpreting their results. You can cite a sufficient existing result directly. Create a new pack only when the receiving use calls for a shipment.
For an independently applicable E.18 flow crossing or A.21 gate, carry its required CrossingBundle or gate anchors into the shipped pack.
Builds on: G.Core defines the shared Part‑G invariants and delegation. Cite the publications or records that define the upstream packs and kits.
Governs: shipping Part‑G outputs as a SoTA‑Pack(Core). A selector-facing outcome gets a selection/parity roster. The kit also exposes applicable PathId/PathSliceId citations and telemetry pins for refresh planning, and may include interop notes used only as citations. An inputs-only pack may omit the roster.
Cites: The definitions of CN‑Spec and CG‑Spec, CHR/CAL, selection, evidence, bridge calibration and refresh orchestration remain with their governing patterns.
G.10:1 - Problem frame — Preserve meaning when shipping
Part G produces publications and records governed by its kits and suites (harvest packs, CHR/CAL packs, evidence graphs, bridge calibration records, log bundles, parity reports). Without a defined pack boundary, shipping can become:
- a tool-specific export without citable contents, or
- a change to legality, defaults or selection semantics made in the course of shipping without being declared, or
- a brittle hand‑off that cannot support RSCR/refresh (no actionable pins/editions/policies attached).
G.10 defines the single normative shipping surface for Part G, SoTA‑Pack(Core), and how to compose a pack that a selector can use and an auditor can cite. G.Core governs the shared invariants.
G.10:2 - Problem — Failures of reuse, legality and refresh
Shipping fails when:
- Format-as-governing-spec. A concrete export format is treated as “the pack,” turning a tool choice into a governing pack definition.
- Editionless hand‑offs. Shipped artefacts omit the edition/policy pins required to replay or compare outcomes, so parity and RSCR become non‑actionable.
- Shipping changes the rules. The pack adds local rules (hidden scalarisation, competing defaults, private gate decisions), fragmenting the governing spec ref.
- Invisible crossings. An actual correspondence, flow crossing or gate lacks the grounds, bundle anchors or policy pins required for its receiving use.
- No method‑of‑obtaining‑output disclosure. Consumers receive outcomes without a minimal, citable trail of which methods, mechanisms, and policies were used, at which editions, to obtain them.
- Refresh orphaning. Telemetry and decay signals exist, but the shipped artefact provides no stable scope keys (
PathId/PathSliceId) and no payload pins for RSCR triggers.
G.10:3 - Forces
| Force | Tension |
|---|---|
| Notation independence | Make packs portable across tools ↔ still make them concrete enough to be used. |
| Completeness vs minimality | Ship enough to be selector‑ready ↔ avoid duplicating governing definition semantics. |
| Continuity vs evolvability | Preserve public IDs across edition bumps ↔ allow legitimate upgrades and deprecations. |
| Cross‑context reuse vs the cost of documenting crossings | Enable reuse across Traditions/contexts ↔ keep crossings explicit and auditable. |
| Telemetry usefulness vs semantic contamination | Export useful signals ↔ avoid turning telemetry into dominance/acceptance without pinned policy. |
| Fast shipping vs refreshability | Ship quickly ↔ ensure RSCR triggers can be planned and scoped (P2W‑path aware). |
G.10:4 - Solution — SoTA‑Pack(Core) as the shipping object and publication kit
A SoTA‑Pack(Core) cites each upstream artefact it includes by stable ID or reference. The pack is independent of notation. It supplies the minimum pins its receiving uses need for selection, audit or refresh. These include applicable path citations and crossing bundles for audit and typed RSCR triggers for refresh.
G.10:4.1 - G.Core linkage (normative)
Builds on: G.Core (Part‑G core invariants; Default Governing Definition Index citation)
GCoreLinkageManifest (G.10) (normative; expands per G.Core:4.2; Nil‑elision applies)
Under Nil‑elision, compute the effective obligations, pins and triggers as union(expand(sets), explicit deltas).
-
CoreConformanceProfileIds:={GCoreConformanceProfileId.PartG.AuthoringBase, GCoreConformanceProfileId.PartG.TriStateGuard, GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted, GCoreConformanceProfileId.PartG.ShippingBoundary} -
RSCRTriggerSetIds:={ GCoreTriggerSetId.RefreshOrchestration }(payload pins:PackId(UTS),publicationScopeId,CNSpecRef.edition,CGSpecRef.edition,PlanItemRefs := ⟨WorkPlanRef, planLocalContentLocator⟩[],AuditPins,UTSRowId[],PathId/PathSliceId, crossing policy pins,TelemetryPinIds, relevant upstream artefact ids) -
DefaultsConsumed:={ DefaultId.PortfolioMode, DefaultId.DominanceRegime, DefaultId.GammaFoldForR_eff }Resolve their definitions throughG.Core.DefaultGoverningDefinitionIndex. -
CorePinSetIds:={ GCorePinSetId.PartG.AuthoringMinimal, GCorePinSetId.PartG.CrossingVisibilityPins } -
CorePinsRequired(pattern delta; pin names only; id‑valued unless noted) := {PackId(UTS),publicationScopeId,contextSliceId?,PlanItemRefs := ⟨WorkPlanRef, planLocalContentLocator⟩[]?(WorkPlanning planned baseline refs),AuditPins(pack‑level pin bundle: edition pins (only on…Ref.edition), policy‑ids, UTS/Path pins; ids only),UTSRowId[],PathId[]?,PathSliceId[]?,CrossingBundleIds := CrossingBundleId[]?,TelemetryPinIds := TelemetryPinId[]?,PortfolioRosterId?,MOOManifestId?(method‑of‑obtaining‑output disclosure; conceptual object id) } -
TriggerAliasMapRef:=∅(canonical trigger ids are used without aliases)
Mode-specific definition pins. Use the applicable
GPatternExtensionblocks inG.10:4.6for additional QD/OEE/interop pins.
G.10:4.2 - SoTA‑Pack(Core) object model (normative; notation‑independent)
SoTA‑Pack(Core) is a shipment object (a pack, not a kit and not a suite) that cites upstream artefacts and exposes pack‑level pins required for downstream use.
SoTA‑Pack(Core) :=
⟨
PackId(UTS),
publicationScopeId,
contextSliceId?,
CG-FrameContext,
entityOfConcern, GroundingHolon?, ReferencePlane, // subject of the pack-level claim, empirical ground when used, and plane
// Governing spec refs (refs + edition pins; semantics governed by their patterns)
CNSpecRef := ⟨A.19 ref, CNSpecRef.edition⟩,
CGSpecRef := ⟨G.0 ref, CGSpecRef.edition⟩,
// Selector-facing selection/parity roster token (conceptual; no formats mandated)
PortfolioRosterId?, // produced by `G.10‑1` as part of composition; may cite ε and the applicable pinned regime/mode refs
// Cited payload packs/kits: resolve each claim’s subject through its own reference. One pack may carry a report on a pump and a comparison of inspection methods tested there.
SoTAHarvestPackId? // e.g., G.2 output id
CHRPackId? // G.3 output id
CALPackId? // G.4 output id
EvidenceGraphId? // G.6 output id
BridgeMatrixId? // G.2/G.7 cited id
BridgeCalibrationTableId? // G.7 output id
SoSLOGBundleId? // G.8 output id
ParityReportId? // G.9 output id
DashboardSliceId? // G.12 output id (optional)
InteropSurfaceId? // G.13 output id (optional)
// Path citation surface (ids only; semantics governed by A.10/G.6)
PathIds := PathId[]?,
PathSliceIds := PathSliceId[]?,
// Planned baseline + audit pins (P2W-aware; ids only)
PlanItemRefs := ⟨WorkPlanRef, planLocalContentLocator⟩[]?,
AuditPins := { id pins… }, // editions only on `…Ref.edition`; includes policies, UTS/Path pins, crossing pins
// Crossing visibility surface (per GateCrossing; ids only)
CrossingBundleIds := CrossingBundleId[]?,
// Telemetry hooks for refresh planning (ids only; PathSlice-keyed; policy-id pinned)
TelemetryPinIds := TelemetryPinId[]?,
// Method-of-obtaining-output (MOO) disclosure (conceptual; ids only)
MOOManifestId?,
Notes?
⟩
PlanItemRefs, when present, resolve the exact U.WorkPlan episteme and locate the baseline content inside it. For A.15.3 content, the locator uses that WorkPlan’s planItemDesignator and any needed rowDesignator; it gives the item or row no independent identity or edition. An A.15.2 baseline stays plan content when it reuses no declaration member. A changed plan must not silently replace the earlier reference carried by the shipped pack.
G.10:4.2.1 - Portfolio roster (normative; pack-governed; governing-definition delegating)
PortfolioRosterId identifies the selector-facing roster. PortfolioRoster@Context records its citations and bindings inside the shipped publication form. It is not a kind of publication face, publication form, interop publication form or carrier.
The roster MUST NOT redefine selection or selected-set semantics (G.5) or parity semantics (G.9). Mode-specific definition pins for QD, OEE or interop come only from G.10:Ext.* blocks.
PortfolioRoster@Context :=
⟨
PortfolioRosterId,
PackId(UTS),
CG-FrameContext,
entityOfConcern,
// Selector operation and default-resolution support
portfolioMode?,
dominanceRegime?,
ε?,
// Published selector outcome and set-result declaration (metadata fields, not local semantics)
selectorOutcomeKind?,
setResultFamily?,
handoffKind?,
subjectKind?,
sourceSetFamily?,
derivedViewKind?,
sourceSetComposition?,
basePaletteRef?,
lensId?,
shortlistId?,
promotionPolicyRef?,
retentionIntent?,
// Selector-facing roster + provenance hooks (ids only)
MethodFamilyRowRefs := MethodFamilyRowRef[]?,
GeneratorFamilyRowRefs := GeneratorFamilyRowRef[]?,
ParityReportId?,
SCRId[]?, DRRId[]?,
// Pin reuse: prefer referencing the enclosing pack’s AuditPins bundle
AuditPins?,
Notes?
⟩
The roster cites G.5 MethodFamilyRowRef = <MethodFamilyId, rowEdition> and GeneratorFamilyRowRef = <GeneratorFamilyId, rowEdition> values. A lineage id alone cannot replay the selected membership or grouping basis; each row ref resolves its immutable source edition.
Shipping these references neither admits the members nor creates their grouping. Preserve the row’s G.5 local or public registry-identity contract. A visible project-local row is not automatically a public registry entry; intentional public registration adds S1/S1′ naming and UTS duties. G.10’s independently applicable pack-shipping requirements and E.24.PUB availability conditions still apply.
Presence rule: PortfolioRosterId MAY be omitted only when the shipped pack is inputs‑only
(e.g., shipping CHR/CAL/evidence without any selector‑consumable selected-set/shortlist output).
The selectorOutcomeKind, setResultFamily, handoffKind, sourceSetFamily, sourceSetComposition, derivedViewKind, basePaletteRef, lensId, and shortlistId fields carry payload metadata or references inside the shipped publication form. They do not define publication face kinds, publication form kinds, interop publication form kinds, or carrier kinds, and the rules of G.5, C.18, C.19 and G.2 remain with those patterns.
Interpretation constraints (normative by delegation). G.Core conformance requires the applicable CN or CG governing specification and its edition to be cited under its own rule. It also enforces the shared invariants for crossing visibility and penalty routing, tri-state guards, set returns, the P2W split, defaults and RSCR trigger types (CC‑G10‑CoreRef).
G.10:4.3 - Shipping procedure (normative)
Compose the pack while preserving the cited definitions:
- S‑1 — Gather & pin. Collect upstream artefact ids and verify the required pins implied by the linkage manifest (edition pins, policy pins, UTS/Path pins).
- S‑2 — Compose
SoTA‑Pack(Core)+ MOO disclosure. Assemble the pack object. For shipped outcomes intended for downstream use, attach aMOOManifestlisting the methods, mechanisms and policies used to obtain them, with the edition of each (ids only; semantics stay with governing definitions). - S‑3 — Publish the selection/parity roster. Give the selector a
PortfolioRosterIdwith the parity and definition pins needed for reproducibility. No format is prescribed. An inputs-only pack may omit the roster under §4.2.1. - S‑4 — Anchor and publish path citations. Ensure A.10 anchors exist and publish/record
PathId/PathSliceIdcitations required for downstream explainability (e.g., theC.23W2AdmissibilityLedger) and maturity rung changes. - S‑5 — Expose CrossingBundle. For each GateCrossing relevant to the shipped artefacts, expose the required
CrossingBundlereferences (fail fast on missing or non‑conformant bundles when required). - S‑6 — Emit telemetry pins for refresh planning. Whenever illumination increases or archive/OEE pin state changes, emit PathSlice‑keyed telemetry with policy‑id and the active
…Ref.editionpins (and QDEmitterPolicyRef/InsertionPolicyRefwhen applicable). - S‑7 — Publish to UTS (twin labels). Mint/refresh UTS Name Cards needed to cite the pack and shipped heads (Tech/Plain twins when required); a claimed cross-context semantic correspondence cites its F.9 relation, separate bounded-use claim and reliance. G.7 calibration and loss notes remain evidence for that use, not identity or permission.
- S‑8 — Optional: ingest interop surface. If
G.13interop is in use, ingest/citeInteropSurface@Contextas annotation-only notes, pinning external index editions; do not redefine interop semantics.
G.10:4.4 - Interfaces & hooks (selector‑ and audit‑facing)
| ID | Interface (conceptual) | Consumes | Produces |
|---|---|---|---|
| G.10‑1 | Compose_SoTA_Pack | G.* outputs, ComparatorSet, Bridges, editions, SCR/DRR deltas | SoTA‑Pack(Core) (UTS row + surfaces) + AuditPins (+ MOOManifestId?) (+ PortfolioRosterId?) |
| G.10‑2 | Publish_UTS | PackId(UTS), UTSRowId[], deprecation/edition‑bump notes | UTS rows/Name Cards for the pack and shipped heads (incl. twins when required) |
| G.10‑3 | Expose_CrossingHooks | GateCrossings, lanes/planes/contexts | CrossingBundle (E.18:CrossingBundle) per GateCrossing; fail on missing/non‑conformant bundles |
| G.10‑4 | Pack_MOO | referenced method/mechanism/policy/edition ids | MOOManifestId (ids only; governing-definition delegating) |
| G.10‑5 | Emit_TelemetryPins | Illumination/archive/OEE events | PathSlice‑keyed telemetry: policy‑id, …Ref.edition (+ QD/OEE pins when applicable) |
| G.10‑6 | Publish_PathCitations | A.10 anchors, PathIds | PathId/PathSlice citations for the C.23 W2 AdmissibilityLedger & rung changes |
| G.10‑7 | Ingest_InteropSurface? | (optional) G.13 InteropSurface@Context | Annotated pack notes citing external‑index editions |
Surfaces remain conceptual per E.5.2; RO‑Crate/ORKG/OpenAlex mappings belong to Annex/Interop and do not affect Core conformance.
Concrete exports and serialisations fall outside this interface set. Their definitions belong to Annex/Interop. Check conformance against G.10’s notation-independent shipping requirements and the applicable G.Core obligations.
G.10:4.5 - Consequence of governing-definition assignment (normative boundary statement)
G.10 alone defines Part G shipping, by delegation to CC‑GCORE‑SKP‑1.
Other G.x patterns may produce artefacts that are shipped, but they must not embed shipping obligations; they cite G.10 shipping surfaces instead.
G.10:4.6 - Extensions (pattern‑scoped; non‑core)
All method‑/generator‑/interop‑specific shipping extension declarations live here as GPatternExtension blocks.
G.10:4.6.1 - GPatternExtension — G.10:Ext.QDArchiveShippingPins
- PatternScopeId:
G.10:Ext.QDArchiveShippingPins - GPatternExtensionId:
QDArchiveShippingPins - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
C.18 - Uses:
{C.18, C.21, G.5, G.8, G.11} - ⊑/⊑⁺:
∅
RequiredPins/EditionPins/PolicyPins (minimum):
DescriptorMapRef.editionDistanceDefRef.edition- active fields from C.21’s DHC replay basis (only when shipped archive telemetry consumes a C.21 DHC coordinate; carry exactly the fields used rather than a generic method-spec or metric-edition pin)
EmitterPolicyRef(policy‑id / ref)InsertionPolicyRef(policy‑id / ref)CharacteristicSpaceRef(id/ref; iff archive partitioning is declared)characteristicSpaceDescriptionRef: U.EpistemeRef(iff partitioning depends on an exact space-definition episteme; resolve it withCharacteristicSpaceRefunder A.19:5.1.1.1)PathSliceId[](to bind telemetry/refresh scope when archive behaviour is present)
RSCRTriggerSetIds: ∅ (covered by G.10 core linkage via GCoreTriggerSetId.RefreshOrchestration)
Notes (shipping-pin discipline):
- When QD archive fields are shipped, this block specifies their required pins.
C.18defines the QD archive semantics used here.
G.10:4.6.2 - GPatternExtension — G.10:Ext.OEEShippingPins
- PatternScopeId:
G.10:Ext.OEEShippingPins - GPatternExtensionId:
OEEShippingPins - GPatternExtensionKind:
GeneratorSpecific - GoverningPatternId:
G.5 - Uses:
{G.5, G.11} - ⊑/⊑⁺:
∅
RequiredPins/EditionPins/PolicyPins (minimum):
TransferRulesRef.editionEnvironmentValidityRegion?(id/ref; iff a region is declared as part of generator-family support)PathSliceId[](scope key for refreshable generator telemetry when present)
RSCRTriggerSetIds: ∅ (covered by the core trigger set)
Notes (shipping-pin discipline):
G.5identifies the generator-family row and its cited open-ended mode. The cited family and transfer rules define that use. The pack carries the pins needed to replay and audit the shipped claim.
G.10:4.6.3 - GPatternExtension — G.10:Ext.InteropCitation
- PatternScopeId:
G.10:Ext.InteropCitation - GPatternExtensionId:
InteropCitation - GPatternExtensionKind:
InteropSpecific - GoverningPatternId:
G.13 - Uses:
{G.13} - ⊑/⊑⁺:
∅
RequiredPins/EditionPins/PolicyPins (minimum):
InteropSurfaceIdExternalIndexRef.editionClaimMapperRef.editionPlaneMapRef.edition?MappingPolicyRef
RSCRTriggerSetIds: ∅ (covered by the core trigger set)
Notes (shipping-pin discipline):
- Record the interop surface that contributed to the pack’s provenance.
G.13governs the crosswalk semantics.
G.10:4.7 - Metadata for shipped results and source views
Ship the metadata the receiver needs to interpret the visible result: subject and source-set kind, the declared surface pins, and any shipped selector outcome’s selectorOutcomeKind, setResultFamily or handoffKind. Use controlled tokens, cited ids or already declared head labels. The roster may carry subjectKind, sourceSetFamily, sourceSetComposition, dominanceRegime, lensId, shortlistId and applicable archive or promotion-policy ids; include those that change interpretation.
For a derived tradition view, retain derivedViewKind, basePaletteRef and its declared qId or reachability/coverage rule. portfolioMode describes selector operation; the declared result kind and source set supply the public label. These fields carry the governing palette, front, archive or selected-set meaning inside the publication. They neither redefine it nor name publication-face, publication-form, interop-form or carrier kinds.
G.10:4.7.1 - Worked publication slice
- If the visible surface is one tradition front under the declared
Q, publishsourceSetFamily=Front,derivedViewKind=TraditionFront, and identify the palette it was derived from withbasePaletteRef=SoTAPaletteDescriptionId. PublishselectorOutcomeKindonly when a G.5 selector outcome is also shipped, andsetResultFamilyonly for itsSetResultOutcomebranch. - If one shortlist is emitted from that derived tradition front, publish
selectorOutcomeKind=SetResultOutcome,setResultFamily=Shortlist,sourceSetFamily=Front,derivedViewKind=TraditionFront,basePaletteRef=SoTAPaletteDescriptionId, and the namedlensIdtogether. - If that same shortlisted surface is emitted as one stable public object, also publish
shortlistId=<...>so a receiver can identify the shortlist by that token. - If one retained tradition archive view is shown, publish
sourceSetFamily=Archive,derivedViewKind=TraditionArchive, and keep the samebasePaletteRefrecoverable. A G.5 selector outcome, when also shipped, carries its admitted outcome kind and, only for aSetResultOutcome, its set-result kind. - If the shortlist is later ordered, publish
setResultFamily=RankedShortlistand keep the declared source set visible. - Use
ChoiceSetonly as a mathematical gloss when the shipped object is explicitly one mathematical analysis artifact rather than the public selected set; it is not asetResultFamilyvalue. - Do not publish
sourceSetFamily=TraditionPalettealone when the visible object is already one derived tradition view; readers need to know which view is on the surface and which base palette it depends on. - Do not publish
TraditionFrontorTraditionArchiveas if they were the default meaning ofTradition. - Do not ask
portfolioModeto tell the reader whether they are seeing one palette, one front, one archive, or one shortlist.
G.10:5 - Consequences
Benefits
- A shipped result becomes selector‑ready and audit‑citable without turning file formats into governing specifications.
- Shipping preserves the semantics of the cited outputs and their governing definitions.
- RSCR/refresh becomes operationally viable because pack‑level scope keys and payload pins are present.
Costs / trade‑offs
- Authors spend time recording the additional pins and publishing the pack surfaces.
- If upstream governing definitions fail to provide citable ids/pins, apply the failure policy in the cited definitions. Shipping will stop or produce a visibly incomplete pack.
G.10:6 - Bias‑Annotation (informative)
- Format bias (Arch/Prag). A popular export format is tempting to treat as “the pack”. Mitigation: keep Core surfaces conceptual (E.5.2); move serialisation recipes to Annex/Interop; keep conformance on semantics.
- Centralisation bias (Gov). A single pattern governing shipping semantics can become a bottleneck.
Mitigation: keep shipping under one governing pattern, but place mode- and method-specific requirements in
G.10:Ext.*extension blocks and cite governing patterns. - Telemetry→dominance bias (Onto/Prag). Promoting telemetry proxies (illumination/coverage) into ranking changes the interpretation of the shipped result. Mitigation: preserve the telemetry/order separation and require a CAL policy‑id for any promotion; record the policy‑id in audit pins/telemetry.
- Interop authority bias (Onto/Epist). External indexes can silently override local legality/typing.
Mitigation:
G.10‑7ingests interop only as cited notes (editions + mapping policy refs), never as a replacement governing spec ref.
G.10:7 - Archetypal grounding
World‑plane (benchmark shipping).
A team working within a CG-Frame ships QD archive results (for example, from MAP-Elites-class or CMA-ME-class methods), a declared selector outcome derived from them, and generator-family results (for example, from POET-class environment generation). The shipped SoTA‑Pack(Core) cites the CHR/CAL packs and records the QD/OEE extension-required pins through the extension blocks so downstream parity and refresh can locate the affected PathSliceIds and reconsider their dependent uses.
Episteme‑plane (synthesis shipping). A team working within a CG‑Frame ships a pluralistic set of admissible methods supported by a literature review and synthesis pack. The shipped pack carries CN/CG spec refs, evidence path citations, and method‑of‑obtaining‑output disclosure; downstream selection uses set‑valued outcomes, and refresh can be scheduled when the synthesis pack or key pins change.
G.10:8 - Conformance checklist (CC‑G10)
For order and illumination, evidence, and bridge or penalty legality, apply the cited governing patterns. The following requirements govern shipping.
| ID | Statement | Verification notes (conceptual) |
|---|---|---|
| CC‑G10‑CoreRef | The pattern satisfies the effective G.Core obligations declared by G.10:4.1 (after profile/set/pin‑set expansion under Nil‑elision). | Check that the linkage manifest is present and that the expanded obligations are not contradicted. |
| CC‑G10.1 (Notation‑independent). | The pack MUST NOT rely on any specific file syntax; cards/tables are conceptual; tool serialisations are informative only. | Look for format‑free conceptual fields; any serialisation is explicitly non‑normative. |
| CC‑G10.2 (Pack parity pins). | If QD archive fields are present, pin DescriptorMapRef.edition, DistanceDefRef.edition, EmitterPolicyRef and InsertionPolicyRef. Include CharacteristicSpaceRef when archive partitioning is declared, with a separate characteristicSpaceDescriptionRef: U.EpistemeRef when the exact defining episteme affects reproducibility; resolve the pair under A.19:5.1.1.1. OEE fields pin TransferRulesRef.edition under their own extension. When a shipped field consumes a C.21 DHC coordinate, carry every active field of that coordinate’s DHCReplayBasis. A non-QD generator use does not acquire QD descriptor or archive pins. | Verify the corresponding G.10:Ext.* block is present and the pins appear in AuditPins and (when relevant) in telemetry pins. |
| CC‑G10.3 (Telemetry discipline). | Any illumination increase or archive edit SHALL log PathSliceId, the active policy‑id, the active values designated by the pinned …Ref fields and the required episteme editions (including OEE TransferRulesRef.edition), and the active EmitterPolicyRef/InsertionPolicyRef when applicable. | Verify emitted telemetry is PathSlice‑keyed and carries the required pins; ensure causes are recorded using canonical trigger kinds (alias labels are optional). |
| CC‑G10.4 (UTS publication & twins). | Shipped heads use UTS Tech/Plain twins under the delegated publication rule. A shipped cross-context semantic-use claim cites its obtaining relation, separate bounded-use proposition and matching reliance; calibration summaries keep their evidence role. Kind and plane claims retain their own governors. | Verify the published names and exact relation/use evidence; a visible CL or policy pin alone is insufficient. |
| CC‑G10.5 (MOO surfaced in shipping). | For every declared selector set-result or archive published, the pack SHALL list the applicable generation/parity method and mechanism ids (e.g., QD EmitterPolicyRef/InsertionPolicyRef, parity harness ids, method refs where the method definition is generative) and the active policy‑id(s) in SCR‑visible bindings and telemetry pins (ids only; governing-definition delegating). | Verify MOOManifestId is present when outcomes are intended for downstream use and does not redefine semantics. |
| CC‑G10.6 (Pack completeness as a citation surface). | The pack cites all included upstream artefacts by id/ref and exposes the required pins (AuditPins, UTS/Path pins, CrossingBundleIds when required). | Verify all present payload artefacts have ids and the pins needed to cite/replay them. |
| CC‑G10.7 (CrossingBundle exposure). | If the shipped artefacts have relevant GateCrossings, expose the required CrossingBundleIds for each under the delegated crossing visibility rule. If there are none, record that fact. Stop shipping if a required bundle is missing or non-conformant. | Check the recorded presence or absence of bundles against the shipped artefacts’ declared crossings. |
| CC‑G10.8 (Baseline binding is explicit when used). | If the shipped pack claims a planned baseline, PlanItemRefs := ⟨WorkPlanRef, planLocalContentLocator⟩[] are present (exact WorkPlan plus plan-local content; no execution logs). | Verify the baseline resolves inside the cited WorkPlan; item and row designators have no independent identity, and decision records or execution logs do not replace the plan. |
| CC‑G10.9 (Extension‑scoped pin declaration). | If QD/OEE/interop fields are present, the corresponding GPatternExtension block is present and its required pins/editions/policies are recorded in AuditPins and in emitted telemetry pins when those pins affect refreshability. | For each mode used, check that its required extension pins are present. |
| CC‑G10.10 (Derived tradition-view shipping). | If the visible shipped surface or shortlisted source is one derived tradition view such as TraditionFront or TraditionArchive, the pack MUST publish the declared sourceSetFamily, keep basePaletteRef=SoTAPaletteDescriptionId recoverable, and carry the derivation basis (derivedViewKind, declared Q, or reachability/coverage rule id) so the reader can distinguish the visible derived view from its base palette. | Verify derived tradition views are shipped as derived views, not as silent redefinitions of the base palette. |
G.10:8.1 - Anti‑patterns and remedies
- AP‑1 Format‑as‑governing‑specification. Remedy: keep Core surfaces conceptual (E.5.2); move serialisation to Annex/Interop; enforce
CC‑G10.1. - AP‑2 Hidden edition drift. Remedy: require
…Ref.editionpins in AuditPins and treat edition changes as RSCR‑relevant via canonical trigger kinds. - AP‑3 “QD archive present” but missing definition pins. Remedy: enforce
CC‑G10.2and theG.10:Ext.QDArchiveShippingPinspin declarations. - AP‑4 Telemetry silently becomes dominance. Remedy: keep telemetry report‑only unless an explicit CAL policy promotes it; require policy‑id recorded (ties to
CC‑G10.3and MOO discipline). - AP-5 Missing required PathSlice key. Remedy: enforce PathSlice‑keyed telemetry and path citations (
G.10‑5,G.10‑6). - AP‑6 Cross-context reuse without its required grounds or anchors. Remedy: cite the actual relation and the receiving-use and reliance grounds required by its governor; include calibration and policy pins when that account requires them. For an independently applicable
E.18flow crossing orA.21gate, expose its required bundle or gate anchors and fail on missing or non-conformant ones (CC‑G10.7,CC‑GCORE‑CROSS‑1). - AP‑7 Interop ingestion rewrites semantics. Remedy: ingest interop as cited notes only; semantics remain in
G.13(G.10‑7,G.10:Ext.InteropCitation). - AP‑8 Derived-view collapse. Remedy: ship
sourceSetFamily,derivedViewKind,basePaletteRef, and the declaredQor reachability basis so the reader can distinguish the derived view from its base palette.
G.10:8.2 - SoTA-Echoing
- Research‑object packaging & provenance. Research-object and RO-Crate approaches package outputs with provenance and version information. G.10 uses that packaging idea as a citation surface for independently governed outputs.
- Reproducibility regimes in ML/AI. Reproducibility checklists, artifact evaluation and benchmark reporting motivate version pins, method disclosure, and separating telemetry summaries from decision criteria unless policy‑promoted.
- Scholarly KG interoperability. For external mappings from systems such as ORKG or OpenAlex,
G.10-7andG.10:Ext.InteropCitationretain index editions and mapping policies as cited notes. FPF legality and typing keep their own governing definitions.
G.10:9 - Relations
Builds on: G.Core; consumes/cites artefacts governed by cited patterns from G.2 (harvest pack), G.3 (CHR pack), G.4 (CAL pack), G.6 (EvidenceGraph), G.7 (bridge calibration), G.8 (SoS‑LOG bundle), G.9 (parity report), optional G.12 (dashboard slice), optional G.13 (interop surface).
Publishes to / used by: UTS (pack identity), selector‑facing consumers (via G.5), audit and assurance publications (SCR/RSCR), refresh orchestration (G.11).
Constrains: tooling exports are downstream; serialisation and repository integration are non‑normative here.
G.10:End
G.11 - Decide Whether and How to Refresh SoTA Packs and Related Results (Telemetry and Decay)
Tag. Architectural pattern (notation-independent)
Status: Stable
Normativity. Normative (unless marked informative)
Stage. run-time and maintenance-time (selective re-computation, republication, and controlled deprecation)
Primary outputs, when their use calls for them. RefreshQueue, RefreshPlan@Context (a local application name for one U.WorkPlan, not a new kind), RefreshReport@Context (record of refresh Work or its audit), DeprecationNotice@Context, and EditionBumpLog@Context. Continued reliance on an already applicable result requires none of these just to certify that refresh was omitted.
Primary hooks. G.Core (RSCR trigger catalogue, alias docking, and Default Governing Definition Index), G.6 (EvidenceGraph; PathId and PathSliceId), G.7 (Bridge Sentinels; CL, Φ, and plane policy pins), G.5 (set-returning selection and dispatch), G.8 (SoS-LOGBundle telemetry hooks), G.9 (parity reruns), G.10 (shipping hooks and pack-level telemetry pins), G.12 (dashboard telemetry pins), B.3.4 (freshness and decay), E.18 (GateCrossing and CrossingBundle visibility), C.18 and C.19 archive, front, and live-pool policy pins, C.23 (SoS-LOG branches and maturity ladders), C.28 (causal-use support results whose SoTA-sensitive fields can change downstream causal-use results).
G.11 cites G.Core for RSCR trigger-kind meaning, CN and CG admissibility, tri-state guards, penalties, set-return semantics, shipping or harvesting delegation, RSCRTriggerKindId values, and default governing definitions.
Refresh plans and reports cite the trigger meanings in G.Core and the governing definitions named in its Default Governing Definition Index.
G.11:0 - Use this when
Use this pattern when a changed premise or condition may alter whether a shipped pack or related result supports its receiving use. The change can concern evidence, source currentness, telemetry, an edition or policy, bridge calibration, a relied-on base relation definition, the named substrate edition or derivation applicability. The affected result can be an evidence set, dashboard, selected set, archive, front, Q-front, term bridge, descriptor set, parity result or use of an A.6.RCD definition.
G.11:0.1 - What goes wrong if missed
A team may rerun unrelated work, continue using an unsupported result or publish a currentness report without repairing the affected use. First determine what the changed basis permits; select and perform upkeep only where it is needed.
G.11:0.2 - What this buys
The practitioner first decides what the available support permits for the receiving use. A shared premise can reopen several dependent results while leaving independent results usable. When upkeep is justified, the refresh plan names that scope, its source and policy basis, and the selected actions. The resulting report records the work performed.
When a later or replacement source may change a claim and the receiving uses must first be found and revalidated, use A.10.1 for the bounded search frame, discovery coverage and gaps, exact-use test, the extent of the claims or actions that may need revision, application of the direct subject guidance, and the independently obtained subject result.
G.11 continues to govern source currentness, decay, refresh planning, and refresh reporting. Use a separately established currentness result or independently obtained subject result to plan later refresh under its governing pattern.
G.11:0.3 - First output
When a structured currentness line is useful for a loop, harness, workflow-store or DPF seed artifact, name its currentness object: for example, the source pack, evaluator, benchmark, harness or workflow edition, pattern seed, PFAD/PFR dependency, selected set, archive, front or publication carrier. The line can state currentness, source decay, edition change, telemetry and scoped refresh refs. Its subject pattern determines whether the artifact improved.
First establish whether the current conditions and available basis already support the receiving use. If they do, retain that result without a new currentness line, WorkPlan, waiver or skip-refresh certificate.
When a later recipient needs a changed limit or retained reason, keep the minimum useful content with the existing result or publication; a RefreshCurrentnessLine@Context can express it when a structured line is useful. Produce a RefreshPlan@Context only for selected planned refresh.
For an underlying claim about a selected set, archive, culture, bridge, evidence, dashboard or shipping, obtain that subject pattern’s result; currentness does not establish its adequacy.
When a currentness record is needed, include the applicable framework edition pins, source packs, publication-carrier currentness, deprecation, supersession and source-decay conditions.
In that record, cite E.4 for the affected framework, E.4.PFR for a framework relation, E.4.PFAD for the framework architecture decision, the exact source contribution and bounded use, including the G.2 pack when that route was used, and E.11 for discovery. For publication, cite E.17 for a source-backed face and return to source and E.24.PUB for the occurrence, form, carrier, audience, bounded use, and availability.
G.11:1 - Problem frame — Keeping shipped SoTA current without global rebuilds
Part G produces selector-ready publications and records for shipment: packs, bundles, evidence graphs, parity reports, and dashboards. Once shipped, they are exposed to:
- telemetry (illumination and archive changes, parity outcomes, dashboard deltas),
- currentness conditions (a relied-on premise changes, a justified review becomes due, or an actual qualification or use window ends),
- edition drift (descriptor, distance, or transfer rules bump; policy pins evolve),
- bridge evolution (CL or plane penalties or calibrations update).
G.11 governs the connection from typed refresh causes and their affected uses to selected plans and execution reports. G.Core supplies cause semantics and the applicable invariants.
G.11:2 - Problem — Why naive refresh breaks comparability and admissibility
A refresh loop fails when any of the following happens:
- Unrelated reruns. Minor edits (e.g., a single Bridge calibration) trigger pack-wide rebuilds without a traceable scope rationale.
- Editionless telemetry. Telemetry signals are recorded without edition pins, making reruns non-comparable and parity-unreplayable.
- Alias-as-semantics. Treating local trigger aliases as definitions of trigger meaning fragments refresh semantics across patterns.
- Silent crossings. Refresh actions implicitly change crossing assumptions (UTS, Path, or policy pins) without a visible CrossingBundle.
- Orchestration smuggles semantics. Refresh introduces new default behaviors (dominance,
PortfolioMode, or Γ-fold) or coerces partial orders into scalars “for convenience.”
G.11:3 - Forces — Minimal recomputation under strict invariants
- Minimal scope vs. completeness. Include every dependent use that can change, while excluding unrelated work. The resulting closure can contain several slices.
- Operational urgency vs. auditability. Actual refresh Work must remain inspectable through the needed pins, references and paths. A currentness decision or an unchanged use must not become a fictitious Work occurrence.
- Alias stability vs. semantic unification. Existing trigger labels must remain usable, but each must resolve by id to its governing trigger definition.
- Modularity vs. orchestration power.
G.11must coordinate harvesting, parity, and shipping without re-implementing them or importing discipline-specific method semantics into core. - Policy-bound behavior vs. “smart defaults.” Ordering, priority and budget handling must follow their declared policies rather than hidden universal rules.
G.11:4 - Solution — RSCR-driven refresh as a P2W-scoped orchestration kit
G.11:4.1 - G.Core linkage (normative)
GCoreLinkageManifest (normative; canonical shape per G.Core; Nil‑elision permitted).
GCoreLinkageManifest := ⟨
CoreConformanceProfileIds := {
GCoreConformanceProfileId.PartG.AuthoringBase,
GCoreConformanceProfileId.PartG.TriStateGuard,
GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted,
GCoreConformanceProfileId.PartG.ShippingBoundary
},
RSCRTriggerSetIds := {GCoreTriggerSetId.RefreshOrchestration},
CorePinSetIds := {
GCorePinSetId.PartG.AuthoringMinimal,
GCorePinSetId.PartG.CrossingVisibilityPins
},
CorePinsRequired := {
RSCRTriggerKindId,
RSCRTriggerAliasId?,
scope: PathSliceId[] | PatternScopeId,
payloadPins{…},
RefreshPlanId?,
RefreshReportId?,
DeprecationNoticeId?,
EditionBumpLogId?,
WorkPlanRef[]?
},
DefaultsConsumed := ∅,
TriggerAliasMapRef := G.Core.TriggerAliasMap.G11
⟩
Expand the profiles, sets and pin sets in the manifest under the G.Core expansion rule. Combine the results with the deltas stated here to obtain the effective conformance ids, trigger kinds and required pins.
TriggerAliasIds (visible; labels only). {G.11:T0…T7} (resolved through TriggerAliasMapRef; G.Core defines the trigger meanings).
G.11:4.2 - Refresh orchestration kit (subject-qualified; conceptual artefacts)
Use the following artifacts for refresh planning and reporting. RefreshPlanId is required on a plan and RefreshReportId on a report. The linkage manifest applies to the triggers and actions that occur; retaining applicable support creates no trigger, plan or report merely to fill their pins.
-
RefreshQueue(conceptual queue). A queue of refresh candidates keyed by affected scope (PathSliceIdrefs preferred where graph-expressed;PatternScopeIdpermitted). One candidate may cover several dependent slices. Ordering, prioritization and batching use declared policies; every queue item carries canonical trigger kind ids. -
RefreshPlan@Context(oneU.WorkPlan). The plan for a refresh is aU.WorkPlanepisteme under A.15.2. It declares intended Work and establishes no performed occurrence. Keep gate decisions outside the plan.RefreshPlan@Contextis this pattern’s name for that plan. The plan declares:RefreshPlanId(UTS-published id; editioned)EntityOfConcernRefandReferencePlanepins (by ref; no implicit widening)TargetScope := PathSliceId[] | PatternScopeId[]PlannedTriggers := RSCRTrigger[](canonical trigger kind ids, scope, and payload pins)PlannedActions := RefreshAction[](each action delegates to a subject pattern)RequiredPins: exact affected source and result editions, policy pins and scope for replayability; UTS pins for public identities used; graph Path pins when the selected scope is graph-expressed or an independently applicable receiving contract requires them.PlannedFillingRows[]?as ClaimGraph content kept inside the WorkPlan under A.15.3 when a value must be pinned against a declaration member defined by its own pattern. A row is addressed only through the WorkPlan and has no separate reference or identity.
-
RefreshReport@Context(record of refresh Work or its audit). An execution or audit report that records:RefreshReportId(UTS-published id; editioned)ExecutedActions[]with independently admitted A.15.1 Work refs and references to their performers identified under A.13, plus links to the resulting artifacts under their subject patterns (for example, a new parity report or pack id). Add F.6 attribution only when the receiving use needs to attribute the Work to its established A.13 assignment.ObservedDeltas(telemetry deltas, admissibility changes, evidence-relation or source-relation changes) as refs and pins, not as untyped proseRSCRRefs[](any RSCR or regression harness artefacts invoked)EmittedNotices[] := DeprecationNoticeId[]andEditionBumpLogId[]- the canonical trigger kinds applied, even when an alias was used
-
DeprecationNotice@ContextandEditionBumpLog@Context. Controlled evolution artefacts that preserve ID-continuity:- DeprecationNotice explains scope, reason class (canonical trigger kind ids), and successor refs.
- EditionBumpLog records edition increments and the pins that justify them.
G.Coregoverns ID continuity and the use of aliases.
G.11:4.2a - Selected-set, archive, and cultural-variant currentness
Use this line when a recipient needs a structured currentness result concerning a selected set, front, Q-front, archive, portfolio lineage, cultural-variant lineage, style or tradition term bridge, path slice, reused A.6.RCD predicate definition, or admitted derived relation kind.
RefreshCurrentnessLine@Context:
governedObjectRef:
currentnessObjectKind:
sourceRecordRef:
editionOrLineagePins:
affectedPathSliceOrScope:
subjectPatternLocator:
receivingUseAndApplicableConditions:
currentnessConclusion:
plannedRefreshAction?:
refreshReportRef?:
currentnessObjectKind may name, for example, a selected set, Front, Q-front, ExplorationArchive, Archive, a cultural lineage, a term bridge, or a reused predicate definition.
Use the line only when its recipient needs this structured currentness result; identify the temporal reference and relevant window within its conditions and pins. plannedRefreshAction? is absent when no refresh is selected, and refreshReportRef? is absent when no such report exists.
The line states applicability for a use, not that the subject claim is true or adequately supported. Use G.5 for selected-set declaration, E.17 and E.24.PUB for publication, C.18 for archive and front relations, C.19 for pool treatment, C.36 for cultural-evolution claims, F.17 for exact local SchemeSenseCells, F.18 for name settlement, F.9 for obtaining Bridges, and A.6.RCD for a derived relation kind.
Use the existing result or publication for a needed currentness conclusion or limitation. Use RefreshPlan@Context, RefreshReport@Context, DeprecationNotice@Context or EditionBumpLog@Context when the corresponding plan, performed work, deprecation or edition change occurs; do not add an empty ticket or notice for unchanged applicability.
When the governed object is a reusable A.6.RCD predicate definition or an admitted derived relation kind, the currentness line pins the exact base definitions, named substrate and edition, authorized derivation operation, and applicability scope. A change to any of them reopens the affected derivation and its dependent uses under A.6.RCD; G.11 schedules the bounded refresh but does not redefine the relation or derivation.
G.11:4.3 - Orchestration semantics (conceptual; delegating to governing definitions)
Use G.11 to plan scoped actions from typed causes; action semantics remain with their subject patterns.
4.3.1 Ingestion. Consume RSCR triggers from:
- telemetry hooks (e.g.,
G.8,G.10,G.12), - freshness and decay events (
B.3.4), - evidence, bridge, policy, edition, relied-on base-definition, named-substrate-edition, or derivation-applicability edits (from the respective subject patterns’ publication faces, forms, or units).
Normalize each ingested signal into an RSCRTrigger with its canonical id, scope and payload pins; retain an alias label when used.
4.3.2 Scope closure over the actual dependencies. Start from the changed premise and the receiving question. Establish the minimal dependency closure that includes every dependent use whose claim, action or required check can change. Minimality excludes unrelated uses. A smaller node count or the first slice found can leave a dependent use out. Follow:
- cited evidence and source relations, using G.6
PathIdandPathSliceIdrefs where those dependencies are graph-expressed; - declared crossings, including G.7 sentinels and required
CrossingBundlevisibility; and - pinned editions and policies consumed by the receiving result.
Recover the exact direct use at each step. A common evidence item, causal assumption, calibration or policy can feed several paths and a result that combines them. Include those branches and the combining use before choosing actions.
Repeated citations to one basis do not make it independent support. G.6 provides the representation and indexes; A.10.1 supplies source-change discovery, direct-use tests and coverage limits when the existing account is insufficient.
Graph reachability alone establishes no dependence. Keep an unresolved branch explicit and continue independently resolved branches.
For a nongraph result, name the exact source, receiving result/use and dependency under PatternScopeId. Use PathSliceId when the scope is graph-expressed or independently required by the receiving contract.
This closure states which conclusions need reconsideration; it neither performs their subject judgement nor commits every affected use to a rerun.
Apply the subject rules with available evidence. An affected branch may retain the same supported conclusion, acquire a narrower use, need refresh or require suspension; an independent result remains reusable on its own qualified basis.
Interpret a B.3.4 trigger for the receiving claim and use. Age alone does not select a disposition. Stop when the present basis gives the warranted answer; keep only a limitation or reason that a later recipient needs. If a subject judgement reveals another consequential dependency, return it to the scope before planning dependent work.
4.3.3 Planning (P2W boundary).
Use C.11.DUA to compare further inquiry with the warranted answer already available. Select an attainable action for what it can change at the receiving horizon, including its full cost, delay and displaced work. C.11 and, where applicable, C.19.2 supply the choice; binding evidence and authority requirements remain in force. Produce RefreshPlan@Context only for the planned actions selected. Possible action forms include:
RerunHarvest(delegates to the selected harvesting or SoTA method, such asG.2; useG.1additionally only when its generator-kit cards or wiring must change)RerunParity(delegates toG.9)RecomputeSelectionOrSetResult(delegates toG.5)RebindBridgeOrCrossing(delegates changes to the obtaining Bridge toF.9, calibration-record changes toG.7, and crossing visibility toE.18and the applicable visibility harnesses)UpdateEvidenceBindings(updates the provenance representation under G.6 after any changed direct relation or subject result has been established under its own governor)ReshipPack(delegates toG.10)UpdateBundle(delegates toG.8)UpdateDashboardSlice(delegates toG.12)EmitDeprecationNoticeorEmitEditionBumpLog(publication units governed by this pattern)
4.3.4 Execution and audit.
When selected actions are performed as Work or Work-bound audit, identify the actual performers and independently admitted occurrences and publish the corresponding RefreshReport@Context.
Assess the returned subject results for each receiving use, including any result that combines refreshed and retained branches. Reuse sufficient existing observations where applicable; a plan supplies no new observation or performance.
When a gate applies, record its G.Core outcome (pass, degrade or abstain) with the policy IDs and cited evidence or source relations.
G.11:4.3a - Causal-use refresh sentinels
When a shipped result consumes C.28, inspect the following changes when they can alter supported or unsupported use, a support-result verdict, its limits or a downstream decision basis:
| Sentinel | Affected result | Refresh pins |
|---|---|---|
| sampling-realizability shift | CounterfactualSamplingRealizabilityResult | target distribution, decision Method and any derivation, physical, ethical, operational, and history constraints, required sampling construction or obstruction, unresolved questions, status, supported use, and unsupported use |
| performed sampling or resulting-data shift | dated sampling Work plus A.10 evidence path and empirical data regime | WorkPlan when used; actual performer identified through A.13; dated Work independently admitted through A.15.1; Method and window; resulting sample or data; provenance and currentness. Add F.6 with the same A.13 assignment only if the refresh decision needs to attribute the Work to that assignment. F.6 identifies neither performer nor assignment. A missing or failed attribution alone leaves independently admitted Work intact. If the finding invalidates a consumed A.13 or A.15.1 premise, recheck that premise and its dependent conclusions. A realizability result cannot substitute for the sampling Work or the resulting data. |
| identification or bound shift | CausalIdentificationResult | data-regime refs, assumptions, identifying derivation, bound or failure witness, sensitivity |
| estimate shift | CausalEstimateResult | identification or design basis, data, estimator Method, diagnostics, uncertainty, sensitivity, and any live estimation-consistency result |
| target-trial practice shift | protocol and mapping results | question/estimand, protocol-to-data mapping, assumptions, estimate/precision, sensitivity and reporting source edition |
| causal-fairness shift | C.28 support result plus D.5 BiasAuditReport@Context | fairness estimand, extra counterfactual-identification assumptions, estimate and consistency result when used, support components and result, affected population, audit limits, and decision |
| causal-representation shift | CausalVariableRepresentationRecord | intervention validity, invariance, abstraction fidelity, query preservation, shift and use limits |
| off-policy or causal-RL shift | OffPolicyCausalEvaluationResult | behaviour/evaluation policies, horizon/history, confounding, overlap, endpoints, estimator and uncertainty |
| simulation-validation shift | simulationResultRef in CausalSupportComponentRefs | model assumptions, validation, sensitivity, supported model use and unsupported realized/interventional use |
| transport-endpoint shift | CausalTransportabilityResult | source/target population, domain, environment and data-generating regime, assumptions, windows, formula and unresolved limits |
These are payload distinctions under existing G.Core trigger kinds. Apply §4.3.2 to every result and receiving use that consumes the changed causal basis, including several results sharing an assumption or data source.
G.11:4.4 - Extensions (pattern-scoped; non-core)
Discipline-specific refresh strategies and generator-specific wiring use GPatternExtension blocks. Queue scheduling uses the policy pins on RefreshQueue and RefreshPlan@Context described in §4.4a. A.15.2 and A.15.3 keep the WorkPlan and its local content separate from dated Work.
G.11:4.4.1 - GPatternExtension — G.11:Ext.TriggerAliases
- PatternScopeId:
G.11:Ext.TriggerAliases - GPatternExtensionId:
TriggerAliases - GPatternExtensionKind:
InteropSpecific(alias docking) - GoverningPatternId:
G.Core - Uses:
{G.Core}(citesG.Core.TriggerAliasMap.G11) ⊑and⊑⁺:∅
Required pins, edition pins, and policy pins (minimum):
RSCRTriggerKindId[](canonical ids recorded on triggers)RSCRTriggerAliasId?(e.g.,G.11:T0…T7as labels only)scope: PathSliceId[] | PatternScopeId
RSCRTriggerKindIds: {RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.FreshnessOrDecayEvent, RSCRTriggerKindId.CrossingBundleEdit, RSCRTriggerKindId.PenaltyPolicyEdit, RSCRTriggerKindId.MaturityRungChange, RSCRTriggerKindId.EvidenceSurfaceEdit}
Notes (wiring-only): Keep the labels T0…T7 and resolve them through G.Core.TriggerAliasMap.G11.
G.11:4.4.2 - GPatternExtension — G.11:Ext.DecayAndDebt
- PatternScopeId:
G.11:Ext.DecayAndDebt - GPatternExtensionId:
DecayAndDebt - GPatternExtensionKind:
DisciplineSpecific - GoverningPatternId:
B.3.4(use-qualified currentness and interpreted planning debt) - Uses:
{B.3.4, G.6} ⊑and⊑⁺:∅
Required pins, edition pins, and policy pins (minimum):
- The receiving claim/use and the changed premise or applicable review condition, with source references where published.
FreshnessWindowDeclRef?,DecayPolicyIdRef?orEpistemicDebtBudgetRef?only when the adopted window, deterioration model or planning measure is used. Their source supplies the meaning; no default expiry or debt budget is required.- The affected dependent claims and uses, expressed as
PatternScopeIdfor nongraph scope orPathSliceId[]for graph scope. Retain graph pins when an independently applicable profile requires them; carrier age alone does not select every use of that carrier.
RSCRTriggerKindIds: {RSCRTriggerKindId.FreshnessOrDecayEvent, RSCRTriggerKindId.EvidenceSurfaceEdit, RSCRTriggerKindId.BaselineBindingEdit}
Notes (wiring-only): B.3.4 determines what the trigger means for the use. Depending on the grounds, continue, narrow or suspend the use, refresh the affected result, or apply an authorized exception. The trigger neither imposes a Refresh/Deprecate/Waive triad nor automatically downgrades the result. Currentness is not assurance of the underlying claim. Budget and priority logic apply only when their interpreted policies are used.
G.11:4.4.3 - GPatternExtension — G.11:Ext.QDRefreshWiring
- PatternScopeId:
G.11:Ext.QDRefreshWiring - GPatternExtensionId:
QDRefreshWiring - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
C.18(QD semantics; descriptor, distance, and insertion) - Uses:
{C.18, C.19, G.5, G.8} ⊑and⊑⁺:∅
Required pins, edition pins, and policy pins (minimum):
DescriptorMapRef.edition,DistanceDefRef.editionCharacteristicSpaceRefwith separatecharacteristicSpaceDescriptionRef: U.EpistemeRef(required when a domain-family coordinate is declared by the QD governing definition; resolve the same-space pair under A.19:5.1.1.1)InsertionPolicyRef,EmitterPolicyRef(policy-bound)- Archive or illumination scope and
policy-idfor each emitted telemetry trigger:PatternScopeIdfor nongraph scope;PathSliceIdwhen graph-expressed or required by an independently applicable parity, evidence or shipping contract.
RSCRTriggerKindIds: {RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.PolicyPinChange}
Notes (wiring-only): Pin the QD inputs so the reruns can be compared. C.18 governs their QD meaning.
G.11:4.4.4 - GPatternExtension — G.11:Ext.OEERefreshWiring
- PatternScopeId:
G.11:Ext.OEERefreshWiring - GPatternExtensionId:
OEERefreshWiring - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
C.19(open-ended exploration and exploration-exploitation logistics) - Uses:
{C.19, G.5, G.8, G.9} ⊑and⊑⁺:∅
Required pins, edition pins, and policy pins (minimum):
TransferRulesRef.edition,EnvironmentValidityRegion(when OEE is declared by the subject patterns)GeneratorFamilyRowRef = <GeneratorFamilyId, rowEdition>andTransferRulesRefwiring pins (as published by G.5 and the governing definitions); resolve the row edition used by the affected result- telemetry scope and
policy-idfor each emitted trigger:PatternScopeIdfor nongraph scope;PathSliceIdwhen graph-expressed or required by an independently applicable parity, evidence or shipping contract.
RSCRTriggerKindIds: {RSCRTriggerKindId.EditionPinChange, RSCRTriggerKindId.TelemetryDelta, RSCRTriggerKindId.PolicyPinChange}
Notes (wiring-only): This module connects OEE refresh triggers to comparable reruns. The governing definition supplies OEE method semantics.
G.11:4.4a - Scheduling and priority policy pins
Scheduling strategies (bandit-style allocation, queueing, cadence policies, early stopping, or manual priority rules) may influence the order and budget of refresh work, but they do not define trigger meaning, action semantics, parity semantics, shipping semantics, or Part-G-wide defaults.
G.11 therefore treats scheduling as policy-bound refresh planning:
RefreshPriorityPolicyIdRefnames the policy used to order or prioritize queue items.BudgetDeclRefnames the time, compute, cost, risk, or cadence boundary for the planned refresh.RSCRTriggerKindId[]comes fromG.Core; scheduling policy does not mint trigger kinds.- the plan for refresh is the
U.WorkPlanlocally calledRefreshPlan@Context; executed refresh is recorded inRefreshReport@Contextor Work-bound audit.
Use an applicable ordinary queue-order rule when it already settles the scheduling choice. If a needed priority or budget rule is missing, resolve that choice under its governing authority before the plan relies on it; missing required policy pins remain a blocker. A heuristic supplies no policy merely by appearing in a queue.
G.11:5 - Archetypal Grounding — System and Episteme (informative)
U.System illustration — Safety-critical maintenance loop (pump and calibration).
A centrifugal pump is serviced under a documented procedure (method description). Sensors report vibration drift (telemetry), and a calibration standard is updated (edition bump). The maintenance team uses G.11 to produce a refresh plan scoped to the affected inspection slices and publishes a refresh report of the executed actions with pins to the updated standard edition and the evidence or source relations. Deprecation notices are issued for obsolete thresholds in the procedure’s acceptance clauses (by subject pattern), preserving ID continuity.
U.Episteme illustration — Living review and benchmark pack (claims and parity).
A claim sheet behind a shipped SoTA pack changes (new evidence, retraction, or revised measurement definition). Calibration evidence changes, potentially affecting a bounded-use claim or a justified loss consequence. The maintainers identify which receiving uses depend on the changed row, retain matching results that remain supported, and use the canonical triggers to plan any needed targeted parity rerun. Re-shipping follows only when the publication needs the resulting change; a CL revision alone grants or withdraws no use.
Paired currentness case. A pack’s export-date label changes, but its relied-on claims, source editions, qualification conditions and receiving use remain unchanged and adequately supported by available information. Retain the result; no refresh plan, waiver or no-refresh notice is needed.
In the paired case, a dependency changes so that the shipped benchmark comparison no longer supports use beyond its stated window. Restrict that comparison and retain the warning with the shipped result so a later receiver cannot infer continued comparability. Plan the targeted check or update when it is justified and obtainable; currentness reporting alone does not repair the comparison.
Three bounded currentness cases. These are constructed applications of the same rule.
- A pump shortlist still consumes the same two immutable method rows, eligibility conditions and source edition. An age alert supplies no changed premise or expired use condition, and the available support remains sufficient for the same triage use. Retain the shortlist; no plan or omission certificate is needed.
- A local selected-set result consumes
PumpReviewBudget-E1. Its replacementPumpReviewBudget-E2changes the allowed review time from 30 to 20 minutes. The existing result cites that budget; no evidence graph is in use. ScopePumpTriageSelectionnames that result, changed budget and dependent eligibility comparison. If re-selection is chosen, plan only that comparison through G.5 underPatternScopeId=PumpTriageSelection, carrying both budget editions and the current task/row refs. Record performed refresh and any required targeted check in the resulting report before republication. - A shipped QD archive expresses its affected dependencies in
PathSliceId=ArchiveCell-Q7. A distance-definition change reopens the comparison using that definition. Retain the slice, old/newDistanceDefRef.edition, descriptor, insertion/emitter and policy pins, and the applicable G.9/G.10 evidence and shipping requirements. An OEE change toTransferRulesRef.editionsimilarly retains its affected graph slice, exact generator row edition and environment-validity scope. Nongraph support elsewhere does not relax these contracts.
Shared calibration, different branch outcomes. Continue G.6:5.5’s constructed case. A correction to calibration qualification K establishes that its uncertainty statement applies only in the low-vibration condition. P1 and P2 both consume K, and selection S uses their acceptance results; P3 uses another calibration. The complete affected scope includes P1, P2 and S, while P3 retains its independent basis.
The inspection team checks the existing configuration records. The records show that M1 was obtained in the qualified condition, so the applicable measurement and acceptance rules preserve its supported result. M2 was obtained outside that condition; the old uncertainty statement cannot support V2 for the proposed use. Preserve the actual measurement history while returning its changed qualification to C.16 and the acceptance question to G.4.
Reconsider S under G.5 because it consumed V2. A permissible shortlist restricted to the supported candidate can answer the immediate triage question; keeping P1 does not preserve the original two-candidate conclusion.
For a later decision that requires V2, the team compares using a qualified independent measurement already available, obtaining a new calibration or measurement, narrowing the planned use and deferring it.
In this case, an available independent measurement covers the required conditions and its assessment fits the decision window; a full repeat campaign would arrive too late and adds no needed coverage. The team selects assessment of that measurement under the subject rules. If it supports V2, S can be reconsidered with both supported branches. If it does not, retain the warranted restricted result. Record a new observation or performed refresh only when it occurs; the plan and the available measurement remain distinct.
G.11:6 - Bias-Annotation (informative)
- Arch bias (toward explicit wiring). Risk: authors feel “over-pinned.” Mitigation: keep the minimum pin set small; push scheduling sophistication into extensions and policies.
- Gov bias (toward audit over speed). Requiring a queue, plan or report for every unchanged result adds paperwork. To avoid it, create queue, plan and report content only for the corresponding need; retain applicable support without an omission certificate, while preserving warnings a later receiver needs.
- Onto and Epist bias (toward one governing definition for trigger meaning). Risk: teams try to localize trigger meaning for convenience. Mitigation: alias docking is allowed, but semantics stay in
G.Core. - Prag bias (toward minimal recomputation). Risk: the first affected slice is mistaken for complete reach. Mitigation: follow the shared basis to every dependent use under §4.3.2 and preserve the coverage limits.
- Did bias (toward readable, reusable artefacts). Risk: an example can hide a condition that matters for another use. Mitigation: recover the receiving question and qualifications of that use before applying the example.
G.11:7 - Conformance Checklist (normative)
| ID | Requirement | Purpose and Notes |
|---|---|---|
| CC‑G11‑CoreRef | A conforming G.11 artefact MUST satisfy the effective core conformance set implied by the GCoreLinkageManifest in G.11:4.1 (profile expansion plus explicit deltas; delegated to G.Core). | G.11 is conformant only if the relevant G.Core invariants and trigger discipline are satisfied. |
| CC‑G11.1 (Slice-scoped planning). | A conforming RefreshPlan@Context SHALL name its complete affected scope as PathSliceId[] (preferred where graph-expressed) or PatternScopeId[] and record canonical RSCRTriggerKindId for each planned cause. Its closure rationale SHALL account for all dependent uses that can change within the declared coverage, including shared-premise branches and combining uses. Pack-wide reruns MAY occur only if that closure spans all slices. | Prevents both unrelated reruns and omission of another dependent use. |
| CC‑G11.2 (Edition discipline; QD and OEE wiring). | When QD, OEE, or both are active, a conforming RefreshPlan@Context and RefreshReport@Context SHALL satisfy the required pin, edition, and policy wiring of the applicable extension blocks: G.11:Ext.QDRefreshWiring, G.11:Ext.OEERefreshWiring, or both. .edition SHALL apply only on …Ref. Missing required pins SHALL block publication. | Keeps replayability strict while keeping method-specific pin lists inside the applicable extension blocks. |
| CC‑G11.3 (Telemetry-metric admissibility). | A refresh SHALL retain the metrics and pins required by the applicable subject method and receiving contract. Any Q, D, QD-score, coverage or regret it publishes SHALL be published as telemetry metrics, and any published IlluminationSummary as a telemetry summary. These values SHALL be excluded from dominance unless a CAL policy explicitly promotes them, and the promoting policy id SHALL be recorded in SCR-visible evidence bindings through the cited subject patterns. | Preserves required telemetry and prevents covert scalarisation. |
| CC‑G11.4 (Bridge penalties). | Any refresh reacting to a sense or kind correspondence change or to a change in an independently governed plane relation SHALL satisfy CC‑GCORE‑PEN‑1. It SHALL preserve the obtaining relation under its direct governor, cited calibration basis and losses, and the separate receiving-use and reliance claims consumed by the refreshed result. It SHALL publish the CL, CL^k, CL^plane values and policy/model pins required by the actual calibration or receiving use, including any required Φ, Ψ and Φ_plane policy ids. Any numerical assurance penalty SHALL satisfy the model applicability and declared-result conditions of CC‑GCORE‑PEN‑1. | Keeps the actual relation, calibration and assurance grounds recoverable during refresh. |
| CC‑G11.5 (Selector invariants). | Any orchestrated re‑selection or selected-set or archive update SHALL (i) satisfy CC‑GCORE‑SET‑1 (delegation), and (ii) cite the selector governing definition (G.5) with the comparator admitted for that use at its applicable edition, and preserve the actual declared outcome: the relevant selected-set kind, narrowed handoff, abstention, or escalation. A changed comparator basis must be explicit under its own governor; G.11 introduces no scalarisation or replacement result semantics. | Prevents refresh from changing order semantics. |
| CC‑G11.6 (Crossing visibility). | Refresh actions that touch cross-context reuse SHALL satisfy CC‑GCORE‑CROSS‑1 under the obtaining relation’s direct governor, retaining its relation reference and the separate receiving-use and reliance grounds required by that governor. An independently governed E.18 flow crossing or A.21 gate that consumes the refreshed content SHALL retain its required harness, pins, lexical constraints and lane checks. Missing required grounds or pins SHALL block the affected publication. | Makes reuse and actual flow crossings or gates checkable under their own requirements. |
| CC‑G11.7 (Use-qualified currentness). | A freshness or decay trigger SHALL be interpreted under B.3.4 for the relied-on claim/use and affected dependencies. Continue on sufficient applicable support without mandatory refresh, deprecation, waiver, WorkPlan or omission certificate. A later receiver SHALL receive the minimum action-changing limitation or reason with the existing result/publication. Publish DeprecationNotice@Context only for actual deprecation; an exception requires actual authority and scope. | Preserves useful currentness warnings without treating age as lost assurance or manufacturing a completion artefact. |
| CC‑G11.8 (No default smuggling). | A conforming G.11 refresh artefact SHALL NOT introduce new defaults for PortfolioMode, dominance, Γ-fold, or guard behavior. If orchestrated steps rely on defaults, the artefact SHALL cite each default’s governing definition through G.Core.DefaultGoverningDefinitionIndex and the applicable subject patterns rather than restating defaults inside G.11. | Protects default definition-citation discipline under orchestration pressure. |
| CC‑G11.9 (Targeted RSCR before republication). | Before changed refresh content is republished downstream, run or cite the required targeted RSCR or regression check for its affected scope. Keep the reference in the existing result/publication or corresponding RefreshReport@Context. Reuse a current matching result for unchanged content; no new execution report is required solely to repeat that reference. A missing required check retains the applicable degrade or abstain outcome under its governing policy. | Keeps actual republication checks while separating their evidence from unnecessary repeated work. |
| CC-G11.10 (Causal-use refresh sentinels). | When a refreshed publication or output consumes C.28, a conforming RefreshPlan@Context SHALL include causal-use sentinel payload distinctions when counterfactual realizability, counterfactual-data identification and bounding, target-trial reporting, causal fairness, causal representation validation, off-policy and causal-RL evaluation, or simulation validation can change supported use, unsupported use, support verdict, assurance, parity, or downstream selection. | Keeps moving causal SoTA from silently invalidating shipped causal-use results while preserving G.Core trigger governance. |
| CC-G11.11 (Relation-derivation dependency refresh). | When a reused A.6.RCD predicate definition or admitted derived relation kind depends on base definitions, a named substrate edition, an authorized derivation operation, or an applicability scope, the refresh plan SHALL pin those dependencies and reopen the affected derivation and dependent uses when one changes. The plan uses existing canonical trigger kinds; it does not mint a relation-specific trigger kind. | Prevents a once-valid derivation from surviving a changed semantic or substrate basis. |
G.11:8 - Common Anti-Patterns and How to Avoid Them (informative)
| Anti-pattern | Symptom | Why it fails | Repair |
|---|---|---|---|
| Full-rerun mania or first-slice stop | An edit reopens every result or only the first located slice. | The scope follows storage or discovery order instead of actual dependence. | Follow §4.3.2 through every affected branch and combining use; retain independent results and coverage limits (CC‑G11.1). |
| Editionless telemetry | Telemetry lacks …Ref.edition | Reruns are non-comparable; parity breaks | Block publication on missing pins (CC‑G11.2) |
| Alias-as-semantics | T* labels are treated as meaning | Trigger meaning fragments; regressions become untestable | Dock aliases through G.Core.TriggerAliasMap.G11; record canonical ids |
| Silent crossing during refresh | Cross-context or plane reuse lacks required relation/use/reliance grounds, or an actual flow crossing or gate lacks required visibility pins. | The reuse or crossing cannot be checked against its governing rule. | Restore the missing grounds; apply E.18/A.21 harnesses to independently governed flow crossings or gates. Block the affected publication when required grounds or pins are missing (CC‑G11.6). |
| Default smuggling | Refresh introduces “helpful” default dominance or PortfolioMode behavior | Competing defaults appear; downstream arguments drift | Cite governing definitions through G.Core.DefaultGoverningDefinitionIndex (CC‑G11.8) |
| Lost currentness warning | A later recipient relies beyond the supported condition or window because the changed limitation was omitted. | The old result can no longer support that receiving use. | Keep the minimum useful warning or decision with the existing result; use a deprecation notice only for actual deprecation. An unchanged immediate use needs no skip-refresh record (CC‑G11.7). |
G.11:9 - Consequences (informative)
- Selective, replayable upkeep. The practitioner can retain a sufficient result or trace selected refresh from its changed premise through dependent uses to performed work and its results.
- Stable semantics with flexible operations. Trigger meaning is centralized (
G.Core), while scheduling sophistication can evolve as policy-bound extensions. - Governing definitions. Refresh coordinates actions under their existing rules. For example,
G.10governs shipping andG.5governs selection. - Cost: pin discipline overhead. Authors carry the ids, editions and policies needed to make refresh comparable and to recover its dependencies.
G.11:10 - Rationale (informative)
G.6 makes a cited basis and its paths recoverable; G.11 connects the affected reach to a supported continuation. A shared premise can require several local judgements, while independent results remain usable. The connection separates three concerns:
- The refresh loop coordinates reruns and republishing;
G.Coregoverns trigger meanings and invariants and identifies the governing definition for each default. - The kit is split across the P2W planning-to-work boundary so that the
U.WorkPlanand its declaration-local planned-filling rows remain planning content while dated Work remains independently established. - Alias stability is maintained by allowing trigger aliases (
T0…T7) while prohibiting them from becoming semantic authorities.
G.11:11 - SoTA-Echoing — Post‑2015 practices aligned (informative)
Each entry follows: claim → practice → source → alignment → adoption status.
0. QD currentness requires visible survey support.
Practice: current QD work is surveyed as approaches, applications, and challenges, with archive, diversity, descriptor, and evaluator-currentness concerns still live.
Source: A survey on Quality-Diversity optimization: Approaches, applications, and challenges, Swarm and Evolutionary Computation 2026, DOI 10.1016/j.swevo.2025.102240.
Alignment: RefreshCurrentnessLine@Context may name selected set, Front, Q-front, ExplorationArchive, Archive, portfolio lineage, descriptor or distance edition, and path-slice scope, while C.18, C.19, and G.5 keep archive, pool, and selected-set meanings.
Adoption: Adopt and bound (survey support changes refresh currentness fields and boundaries; it is not the governing ontology source).
0a. Open-ended engineering outputs need source and evaluator currentness.
Practice: self-improving-agent, AlphaEvolve-style, and DeepEvolve-style lines use generated variants, external knowledge, evaluators, tests, archives, and empirical validation.
Source: Darwin Godel Machine arXiv:2505.22954, AlphaEvolve arXiv:2506.13131, and DeepEvolve-style deep-research augmentation arXiv:2510.06056.
Alignment: G.11 refresh records carry source, evaluator, descriptor, policy, edition, lineage, and report refs; generated method text, evaluator success, and archive update keep their subject patterns.
Adoption: Adopt and adapt (refresh tracks currentness and complete affected scope; generated text, gate passage and performed work retain their different subject claims).
-
Deployed evaluation needs monitoring and justified refresh conditions. Practice: production ML teams monitor data and model behaviour, relate staleness to prediction quality, and use observed changes to decide when to update. Source: Breck et al., The ML Test Score (IEEE Big Data, 2017), Model 4 and Monitor 4; Amershi et al., Software Engineering for Machine Learning: A Case Study (ICSE-SEIP 2019), §II.B. Alignment: G.11 uses typed causes, edition and policy pins, and the receiving-use comparison in §4.3 to select a replayable response. Adoption: Adopt and adapt the monitoring and update-decision practices to affected-use refresh; retain sufficient applicable support.
-
Non-stationarity requires explicit drift and decay handling, not ad-hoc updates. Practice: continual learning emphasizes non-stationarity as a first-class maintenance condition. Source: Parisi et al., Continual Lifelong Learning with Neural Networks (
arXiv:1802.07569, 2019); De Lange et al., A Continual Learning Survey (arXiv:1909.08383, 2021). Alignment:B.3.4supplies use-qualified currentness.G.11interprets changed conditions and justified review signals before planning an affected-scope refresh; elapsed time alone implies neither truth decay nor deprecation. Adoption: Adapt (refresh of conceptual artefacts and evidence closures, not untracked model mutation). -
Quality-Diversity requires archive semantics and comparability under descriptor and distance evolution. Practice: QD methods treat the archive as the primary result and track changes under policy and edition conditions. Source: contemporary QD families such as CMA-MAE (
arXiv:2205.10752) and differentiable QD (arXiv:2106.03894). Alignment: QD-specific meaning lives with the subject patterns;G.11:Ext.QDRefreshWiringensures edition pins and scope pins exist so targeted archive refresh is admissible. Adoption: Adopt (set and archive preservation; no covert scalarization). -
Open-endedness co-evolves environments and agents; transfer rules must be versioned. Practice: POET-class open-ended systems require explicit transfer rules and environment validity constraints. Source: Wang et al., POET (
arXiv:1901.01753, 2019); later generator-family claims require a namedG.2SoTA pack or exact current source. Alignment:G.11:Ext.OEERefreshWiringrequiresTransferRulesRef.editionand scope pins so refresh reruns remain comparable and auditable. Adoption: Adopt and adapt (adapted to Part G pin and UTS publication discipline). -
Efficient orchestration benefits from bandit and early-stopping scheduling, but scheduling must not redefine trigger, action, parity, shipping, or Part-G-wide default semantics. Practice: modern hyperparameter and experiment scheduling uses bandit-style resource allocation and asynchronous early stopping. Source: ASHA (
arXiv:1810.05934) and BOHB (arXiv:1807.01774) as representative post-2015 scheduling practice. Alignment: scheduling is expressed asRefreshQueueandRefreshPlan@Contextpolicy pins (RefreshPriorityPolicyIdRef,BudgetDeclRef) so core semantics remain stable and theU.WorkPlanstays separate from dated Work. Adoption: Adapt (useful practice, but quarantined outside core norms).
G.11:12 - Relations
Builds on: G.Core (Part‑G invariants; RSCR trigger catalogue; alias docking; Default Governing Definition Index), G.6 (EvidenceGraph, PathId and PathSliceId), G.7 (Bridge sentinels; CL, Φ, and plane pins), G.5 (selector and set-return), G.8 (bundle telemetry hooks), G.9 (parity), G.10 (shipping hooks), B.3.4 (freshness and decay), E.18 (GateCrossing visibility).
Coordinates with: A.10.1 for source-change discovery, direct-use tests and affected reach; C.11.DUA for choosing worthwhile inquiry or a sufficient present answer; G.12 for dashboard telemetry pins; A.6.RCD for reopening reused predicate definitions and derived relation kinds when their base definitions, named substrate edition, authorized derivation operation or applicability changes; C.18 and C.19 for archive, front and live-pool policy pins; C.32.P2S when telemetry, decay or freshness reopens architecture problem-to-structure carry-through; C.23 for SoS-LOG branches and maturity ladders; C.28 for causal-use support results, verdicts, supported and unsupported uses, and causal-use sentinel payloads; F.15 for RSCR harness publications when present.
Publishes to: UTS (refresh plan, refresh report, deprecations, edition bumps), and to the relevant subject patterns’ publication faces, forms, or units through delegated actions.
G.11:End
G.12 — DHC Dashboards (Discipline-Health Time Series and Views)
Status: Stable
G.12:0 — Use This When
Use G.12 when a team needs several recorded C.21 coordinate results arranged across windows, a dashboard view over them, or refresh wiring for that view.
Start from the C.21 results. State the discipline, intended use, ClaimScope, coordinate-result refs, and windows before choosing a layout. A local view is sufficient when there is no audience publication or refresh use.
Do not use G.12 for one ordinary field-health claim, to manufacture measurements from rows, to turn a dashboard into evidence or authority, or to require publication and telemetry for every C.21 use.
G.12:1 — Intent
Produce a reproducible discipline-health series and view while keeping these objects separate:
- C.16 measurement results and their C.2.1 coordinate-result epistemes;
- one optional C.2.1
DHCSeriesepisteme that orders the coordinate-result refs by window; - rows and slices that represent those results or series;
- any E.24.PUB publication occurrence, form, carrier, audience, and availability interval; and
- any measurement, series-assembly, rendering, upload, or refresh Work.
G.12:2 — Problem Frame
Dashboards drift or become misleading when they:
- treat
ClaimScopeand a selectedTargetSliceas one field; - copy a value without its C.21 replay basis;
- average nominal or ordinal values or mix Units;
- hide normalization, distance, comparison, or target-band rules;
- require a Bridge for every source difference or omit F.9 when distinct local senses are related;
- turn a row, screenshot, UTS name, form, or carrier into the measurement or series episteme;
- turn selected sets or archives into one scalar winner; or
- rebuild everything because changed definition and evidence pins cannot be localized.
G.12:3 — Forces
| Force | Tension |
|---|---|
| Readable view vs replay | A useful dashboard should be easy to read, while every relied-on coordinate must return to its exact definition and result. |
| Stable history vs changed definitions | A new method or Scale edition may invalidate trend comparability without changing historical results. |
| Optional publication vs local use | A local view may be enough; audience availability adds a separate publication relation. |
| Selective refresh vs process burden | Refresh needs pins that identify the changed basis and the affected results, but a one-off view needs no telemetry framework. |
| Set-valued results vs headline pressure | A view can summarize without manufacturing a scalar winner. |
G.12:4 — Solution
G.12:4.0 — G.Core linkage
This pattern consumes G.Core obligations only for the branches opened.
GCoreLinkageManifest (G.12)
CoreConformanceProfileIds:={GCoreConformanceProfileId.PartG.AuthoringBase, GCoreConformanceProfileId.PartG.TriStateGuard, GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted, GCoreConformanceProfileId.PartG.ShippingBoundary}.RSCRTriggerSetIds:={GCoreTriggerSetId.BridgeCalibrationKit}only when crossing or refresh wiring is used.RSCRTriggerKindIds:={RSCRTriggerKindId.LegalitySurfaceEdit}only when a persisted series or view depends on that surface. Optional panels add only their own declared trigger kinds.DefaultsConsumed:=∅; portfolio defaults become current only throughG.12:Ext.PortfolioTelemetry.CorePinSetIds:={GCorePinSetId.PartG.AuthoringMinimal, GCorePinSetId.PartG.CrossingVisibilityPins}with nil-elision.
The minimal durable series basis is DHCSeriesRef.edition, DisciplineRef, IntendedUse, ClaimScopeRef, exact coordinate-result refs, and their windows. Each coordinate resolves the complete DHCReplayBasis from C.21. PathSliceId[], crossing pins, public-name rows, shipping pins, publication refs, and telemetry pins are conditional.
G.12:4.1 — Objects
| Local name | Exact object | Boundary |
|---|---|---|
DHCCoordinateResultRef | Ref to one persisted C.21/C.16 coordinate-result episteme and its active C.21 replay basis. | It is not a row, series, evidence path, or acceptance decision. |
DHCSeries | One C.2.1 episteme whose EntityOfConcern is the discipline and whose ClaimGraph orders coordinate-result refs by their stated windows under one intended use, ClaimScope, and comparison basis. | It is not a public U-kind, publication occurrence, dashboard, carrier, or Work. |
DHCRow | One representation element showing the reference to the coordinate result and selected readable fields. | It does not compute, establish, or replace the result. |
DashboardSlice | A dashboard representation or grouping over references to the rows, results or series. Use C.29 when an unresolved mathematical representation or transfer question matters to the dashboard use. | It adds no comparison, normalization, acceptance, or selection semantics. |
DHCTelemetryPin | A G.11-facing refresh payload with a canonical trigger, exact affected scope, and changed definition, window, evidence, or policy pins. | It is not evidence, currentness, an edition relation, or refresh Work. |
| dashboard publication | An E.24.PUB occurrence for one exact selected episteme edition expressed by the dashboard form, with its audience, bounded use, carrier, and availability interval. The selected edition may be a series, a coordinate result, or a view independently constituted as a C.2.1 episteme. | A view element, UTS row, rendering, upload, or release label does not supply episteme identity or make publication obtain. |
Conceptual forms:
DHCSeries := <
DHCSeriesRef.edition,
DisciplineRef,
IntendedUse,
ClaimScopeRef,
ComparisonBasis,
CoordinateResultRefs[],
WindowOrder,
DHCDefinitionSetRef.edition?,
TargetSliceRef?,
CurrentnessRuleRef?
>
DHCRow := <
RowId,
DHCCoordinateResultRef,
Window,
DisplayedValue,
DisplayedScaleOrUnit,
DisplayedStance?,
DisplayAnnotations?
>
DashboardSlice := <
DashboardSliceId,
DHCSeriesRef.edition?,
IncludedCoordinateResultRefs[],
IncludedRowIds[],
ViewSpecRef?,
Annotations?
>
TargetSliceRef is present only when the series construction or publication consumes an A.2.6 selection. The ClaimGraph must then state how each selected slice belongs to or is covered by the authoritative ClaimScope. Identify changes to the time window. Do not silently substitute “latest.”
G.12:4.2 — Method of obtaining the result
Stage A — Select what the view is about
- Select the coordinate results. Use persisted C.21 coordinate-result refs for one already identified discipline.
- Fix use, scope, and windows. Name IntendedUse and ClaimScope. Add a
TargetSliceRefonly when the computation or publication consumes it, and state its relation to the scope. - Check replay identity. For every coordinate, resolve the C.21
DHCReplayBasis: Characteristic, Scale, Unit when current,DHCMethodRef.edition, the Method used, MethodDescription edition when used, model or calibration pins when used, time or population basis, and any active distance or definition-set edition. - Choose the comparison branch. Directly comparable C.16 readings need no Bridge. When the comparison uses a relation between distinct local senses, cite the obtaining F.9 relation, observed loss, and a separate affirmative bounded-use claim naming direction, use rule, and loss tolerance, with the current supporting reliance required by F.9. Add reference-plane routing only when the use crosses reference planes; cite its exact basis, and assess any numerical assurance consequence under the applicable B.3 model for the named use.
- Choose panels for the receiving use. Portfolio, QD, open-ended, maturity, SoTA, shipping, and advanced-view fields appear only through their extension blocks.
Stage B — Construct or update content
- When new coordinates are required, obtain their C.16 measurement results and coordinate-result epistemes with the active C.21 replay basis. Identify the Method used and the dated measurement Work, and the MethodDescription, model, and calibration basis required for that measurement use. Keep these objects separate. G.12 creates none of them from a dashboard row.
- When the receiving use needs one series episteme, assemble or revise the
DHCSeriesClaimGraph from exact coordinate-result refs and windows. If a series episteme is unnecessary, construct the local view over the result refs. Series assembly may be dated Work; its resulting episteme remains separate from the Work or work record. - Apply A.18 and the A.19/G.0 rules used for comparison, normalization, distance or aggregation. Nominal and ordinal values remain non-arithmetic unless an explicit lawful transformation creates another Scale.
- Construct the
DHCRowandDashboardSlicerepresentations. Keep each displayed claim linked to its result and replay basis. Label a reported rate with the population it actually describes. If the user needs a decision about a wider population, show the range or estimate supported for that population and whether it settles the decision. A filter that selects fewer rows can change the question a panel answers.
Stage C — Publish or refresh only when required
- If public designators are needed, use F.18 for names of already constituted series or views. A name row is not publication.
- If an audience must be able to obtain a selected episteme edition, establish E.24.PUB with that exact edition, audience, bounded use, dashboard form, carrier, and interval. If the view itself is the selected episteme, recover its independent C.2.1 identity. Otherwise, if the view presents several source epistemes, identify each selected edition and its publication occurrence. Construct a series episteme only when the receiving use needs that additional ordered account.
- If changed definitions, windows, evidence paths, crossing bases, or policies must trigger selective maintenance, emit G.11 telemetry pins naming the affected result or series slice. Otherwise stop without refresh wiring.
G.12:4.9 — Optional Extensions
An extension adds only the panel-specific fields, pins, and triggers consumed by that view. It does not redefine C.21, C.16, comparison, evidence, publication, selection, or refresh semantics.
G.12:4.9.1 - G.12:Ext.SoTAPalette — SoTA palette alignment
PatternScopeId:G.12:Ext.SoTAPaletteGPatternExtensionKind:InteropSpecificGoverningPatternId:G.2- Optional pins:
SoTA_PackRef.edition?, exact F.17 cell refs, and obtaining F.9 relation refs when alignment is displayed. - No additional trigger kind by default.
G.12:4.9.2 - G.12:Ext.PortfolioTelemetry — selector result panel
PatternScopeId:G.12:Ext.PortfolioTelemetryGPatternExtensionKind:MethodSpecificGoverningPatternId:G.5- Conditional values:
TaskSignatureRef?, resolvedDominanceRegime, resolvedPortfolioMode, and exact selector result and basis refs. - Set-returning semantics remain visible. A scalar headline is only a view annotation unless a separate policy lawfully constructs it.
G.12:4.9.3 - G.12:Ext.QDTelemetry — illumination or archive panel
PatternScopeId:G.12:Ext.QDTelemetryGPatternExtensionKind:MethodSpecificGoverningPatternId:C.18- Conditional pins:
DescriptorMapRef.edition,DistanceDefRef.edition,CharacteristicSpaceReffor the selected space and separatecharacteristicSpaceDescriptionRef: U.EpistemeRefwhen an exact description is consumed (resolve the pair under A.19:5.1.1.1),InsertionPolicyRef,EmitterPolicyRef?,ArchiveSnapshotRef?, andPathSliceId[]when refresh uses them. - Illumination and coverage stay telemetry unless a separate accepted policy promotes them into the comparator, dominance set, or selected-set criteria under C.18’s trade-off and authority conditions.
G.12:4.9.4 - G.12:Ext.OpenEndedTelemetry — open-ended or transfer panel
PatternScopeId:G.12:Ext.OpenEndedTelemetryGPatternExtensionKind:GeneratorSpecificGoverningPatternId:C.19- Conditional pins:
TransferRulesRef.edition,EnvironmentValidityRegionId?,ProbeBudgetPolicyId?, andPathSliceId[]. - Open-ended signals do not become dominance objectives by display.
G.12:4.9.5 - G.12:Ext.MaturityLadderPanel — maturity view
PatternScopeId:G.12:Ext.MaturityLadderPanelGPatternExtensionKind:DisciplineSpecificGoverningPatternId:G.8- Conditional values:
MaturityCardRef,MaturityRungId?, and evidence-path refs when the displayed rung relies on them. - Adds
RSCRTriggerKindId.MaturityRungChangeonly for a refresh-wired view.
G.12:4.9.6 - G.12:Ext.PackInclusion — shipping stub
PatternScopeId:G.12:Ext.PackInclusionGPatternExtensionKind:InteropSpecificGoverningPatternId:G.10- Conditional values: exact pack ref, selected
DHCSeriesRef.editionorDashboardSliceRef, and the replay or shipping pins required by the included claims. - G.10 governs shipping; this extension only identifies what is included.
G.12:4.9.7 - G.12:Ext.ViewFamilySeed — specialized views
An embedding, prediction, change-point, or drift panel needs its selected method, inputs, limitations, and policy before it can affect a claim or decision.
G.12:5 — Interfaces
| Interface | Consumes | Produces |
|---|---|---|
Create_DHCSeries | exact coordinate-result refs, discipline, intended use, ClaimScope, windows, comparison basis, optional definition-set and target-slice refs | one C.2.1 DHCSeries episteme edition |
Update_DHCSeries | prior series edition, added or replaced exact result refs, affected windows, edition rule | successor series episteme edition only when C.2.1’s historical-continuation predicate holds, plus exact edition relation when asserted |
Render_DHCView | exact result or series refs, view specification, annotations | DHCRow[] and/or DashboardSlice representations |
Publish_DHCView | exact selected episteme edition (series, coordinate result, or independently constituted view episteme), dashboard form, and E.24.PUB audience, bounded use, carrier, and interval | obtaining publication relation for that edition when its predicate holds |
Emit_DHCTelemetry | exact changed definition, window, evidence, crossing, or policy pin and affected slice | G.11-facing telemetry payload |
| optional panel interfaces | the values specified by the corresponding extension | only that panel’s representation and conditional refresh pins |
G.12:6 — Conformance Checklist
| Check | Passing condition |
|---|---|
CC-G12-1 | Every displayed coordinate resolves one exact C.21/C.16 result episteme and the active C.21 replay basis. |
CC-G12-2 | ClaimScope is authoritative; TargetSlice is optional, consumed explicitly, and related to that scope. |
CC-G12-3 | Direct same-semantics comparison uses C.16 conditions without a Bridge; a comparison using a relation between distinct local senses cites the obtaining F.9 relation, observed loss, and a separate affirmative bounded-use claim with its direction, use rule, loss tolerance, and the current supporting reliance required by F.9. |
CC-G12-4 | Characteristic, Scale, Unit, Method, MethodDescription, model, calibration, Work, result, result episteme, series episteme, row, slice, publication, form, and carrier are not collapsed. |
CC-G12-5 | Numeric, ordinal, target-band, normalization, distance, comparison, and aggregation operations cite their exact lawful definitions. |
CC-G12-6 | A series ClaimGraph identifies exact result refs, windows, intended use, ClaimScope, and comparison basis; content change uses the applicable edition rule. |
CC-G12-7 | Rows and slices are view-only representations. They introduce no new objective, scalar winner, evidence, acceptance, or authority. |
CC-G12-8 | Public names and E.24.PUB publication are conditional and separate; local dashboards need neither. |
CC-G12-9 | Refresh telemetry appears only for a named maintenance receiver and identifies the exact affected slice and changed pins. |
CC-G12-10 | Optional panel fields appear only with their extension and preserve the source pattern’s set, archive, maturity, transfer, shipping, or palette semantics. |
CC-G12-11 | The effective G.Core obligations are expanded by value; nil-elided or unused branches are not made mandatory. |
G.12:7 — Bias-Annotation
G.12 counters screen-first, “latest”-by-default, scalar-winner, and publication-as-truth bias. An uncluttered view can hide incompatible definitions, while a dense technical record can make a simple trend unreadable. Start from exact result claims, show the smallest useful view, and give readers a way to reach replay and refresh details without crowding the view.
G.12:8 — Consequences
Benefits. Dashboard claims can be read quickly and replayed from exact result and definition refs. Historical results remain distinct from changed definitions, and refresh can be local when needed.
Costs. A relied-on trend needs exact result, scope, window, and replay identities. Publication and refresh add their own conditional work.
G.12:9 — Relations
Builds on: C.21, C.16, C.2.1, A.2.6, C.29, E.24.PUB, and G.Core only for the active Part-G branches.
Coordinates with: G.6 and G.11 for relied-on evidence and refresh; A.19 and G.0 for comparison or aggregation; F.9 when a comparison uses a relation between distinct local senses; F.18 for optional public names; G.5, C.18, C.19, G.8, G.10, and G.2 through the declared extensions.
G.12:10 — Author’s Quick Checklist
- Name the discipline, intended use, ClaimScope, exact coordinate-result refs, and windows.
- Resolve the C.21 replay basis for every coordinate.
- Add TargetSlice only when consumed and state its relation to ClaimScope.
- Use the direct comparison branch unless use across distinct local senses requires F.9.
- Keep series episteme, measurement Work/result, row, slice, publication occurrence, form, and carrier separate.
- Add public naming, publication, evidence paths, assurance, optional panels, and refresh only for named receivers.
G.12:11 — Worked Micro-examples
Decision-making dashboard. A local view shows exact ReproducibilityRate, FormalRecognitionStatus, PracticeAdoptionRate, AlignmentDensity, TraditionShareEntropy, and TraditionShareConcentration result refs. A row showing 60% adoption among respondents gives the number who replied, the number eligible and the period. The portfolio owner asks whether at least half of all eligible units adopted the practice. Show the C.21 result for that population. If every admissible count reaches the threshold, it is met. If every count falls short, it is missed. Two admissible counts on opposite sides leave the answer open. If only an outer bound crosses the threshold, report the limit of that calculation. A tighter calculation with the existing observations may settle the question. The owner decides what to do. Entropy and HHI occupy separate rows with their directions visible. A portfolio panel preserves the G.5 selected set. No audience publication is claimed.
Evolutionary architecture dashboard. A DHCSeries episteme orders exact reproducibility and DisruptionBalance results over declared windows. An optional open-ended panel shows transfer events as telemetry. A later E.24.PUB occurrence makes the selected series edition available to a named audience in dashboard form; that occurrence neither changes the series content nor turns the rendering Work into the health result.
G.12:End
G.13 - External Interop Hooks for SoTA Discipline Packs (conceptual)
Use this when. A SoTA author or downstream consumer needs repeatable mappings from an external scholarly source into FPF inputs, with the source snapshot, transformation and receiving use recoverable. Start by naming that source and the input needed by the receiving operation. A sufficient direct source citation needs no interop kit.
Use :4.3 to establish the mapping from source meanings and actual records, then publish its inputs, outputs, conditions and revisions. Provider-specific ingestion, embedding and entity-resolution methods retain their own grounds; the extensions below identify their required wiring.
Status. Stable
Normativity. Normative when a G.13 surface is authored, emitted or consumed. Informative otherwise.
G.13:1 - Problem frame
External scholarly indexes and discipline repositories supply inputs for SoTA authoring. Examples include concept or dataset registries, paper/claim graphs, taxonomy stores and science-of-science indicator feeds. Their values must retain their source meaning and limits when used in FPF characterization, comparison, selection or reporting.
If the mapping assumptions are unstated, a receiver can miss an edition change or a cross-plane or cross-context use. A receiver may also take an external signal as the governing specification for the receiving claim.
G.13 provides registration, alignment and telemetry hooks for whichever Part-G use needs those sources.
A source may supply G.2 synthesis, G.3 characterization, G.5 selection, G.9 comparison or G.12 reporting inputs. G.10 carries an interop surface when shipping is needed; G.11 reconsiders dependent uses after a relevant change. Choose the uses the receiving task requires.
G.13:2 - Problem
External sources publish claim‑adjacent signals (citations, concept graphs, “task/method” tags, replication links, dataset usage, disruption‑style indicators, benchmark metadata). These signals can support audit and the generation of new inputs, for example by helping to build a palette, explore a declared set result or discover a candidate bridge. The mapping can fail in the following ways:
- CN/CG spec-ref leakage. A receiver treats external numeric signals as lawful scores without showing how they are bound to the applicable CHR/CAL/CG definitions.
- Implicit crossings. An opaque transformation hides the actual correspondence, its receiving-use grounds or the bundle anchors required by a real flow crossing or gate.
- Edition drift + refresh brittleness. Snapshots change, schemas drift, indicator definitions get revised; without edition‑pinned interop surfaces and typed trigger causes, parity and dashboard stability degrade.
- Evidence disconnect. A derived feature without recoverable source and transformation grounds cannot be checked or repaired in the receiving use.
- Format‑as‑norm. A convenient serialisation (KG export, JSON schema, RO‑Crate, etc.) is taken as the specification, undermining notation independence.
G.13:3 - Forces
| Force | Tension |
|---|---|
| Notation independence | Useful serialisations vs the requirement that conformance is judged on conceptual surfaces. |
| Pluralism vs parity | Diverse scholarly traditions and indexes vs lawful, edition‑aware comparability and reproducibility. |
| Interop as generation input | Interop should speed SoTA authoring. Using it only to illustrate an audit report misses that use. |
| Planes & bridges | Cross‑plane/context reuse must remain explicit and auditable rather than implicit in “aligners”. |
| Telemetry vs dominance | External telemetry should inform exploration and refresh without silently changing selector semantics. |
| Operational drift | Source editions change; typed causes and payload pins must identify the affected inputs and receiving uses. |
G.13:4 - Solution — Conceptual interop kit: registered sources, alignment cards, feature derivations, and RSCR‑wired telemetry
G.13:4.1 - G.Core linkage (normative)
Builds on: G.Core.
GCoreLinkageManifest (normative).
(Canonical form, Nil‑elision, and Expansion rule are defined in G.Core.)
GCoreLinkageManifest := ⟨
CoreConformanceProfileIds := {
GCoreConformanceProfileId.PartG.AuthoringBase,
GCoreConformanceProfileId.PartG.UTSWhenPublicIdsMinted,
GCoreConformanceProfileId.PartG.ShippingBoundary
},
RSCRTriggerSetIds := {GCoreTriggerSetId.SoTAHarvestSynthesis},
RSCRTriggerKindIds := {RSCRTriggerKindId.BaselineBindingEdit}, // delta: planned‑baseline linkage edits can be interop‑relevant
CorePinSetIds := {
GCorePinSetId.PartG.AuthoringMinimal,
GCorePinSetId.PartG.CrossingVisibilityPins
},
CorePinsRequired := {
// Interop pins (G.13‑specific; avoid duplicating `GCorePinSetId.PartG.CrossingVisibilityPins`)
ExternalIndexRef.edition,
ClaimMapperRef.edition?,
MappingPolicyRef?,
PlaneMapRef.edition?,
ScaleEmbeddingSpecRef.edition?,
EvidenceGraphId?,
InteropSurfaceId?
},
DefaultsConsumed := {DefaultId.PortfolioMode, DefaultId.DominanceRegime}
⟩
Payload‑pin note (informative). When emitting RSCR triggers for interop‑driven changes, payload pins should include the edited edition/policy identifiers, the impacted scope, and the applicable crossing‑visibility pins (per GCorePinSetId.PartG.CrossingVisibilityPins) when crossings/UTS/paths are involved.
G.13:4.2 - Interop kit objects & surfaces (pattern-governed; notation‑independent)
All objects below are conceptual. Any concrete serialisation belongs to Annex/Interop or tooling notes and is not normative for Part‑G conformance.
-
ExternalIndexCard@Context— registration of an external source and its snapshot.Shape (conceptual):
⟨ ExternalIndexId, ProviderName?, ExternalIndexType, CoverageScope, Licence?, ExternalEdition, FreshnessWindow?, entityOfConcern, GroundingHolon?, ReferencePlane, Notes? ⟩Intent. Register the provider snapshot as
ExternalEditionon a stable, citable source card. Downstream artefacts cite the card edition throughExternalIndexRef.edition; they do not copy the provider snapshot id into other cards. -
ClaimMapperCard@Context— a declaration of the recipe whose application maps an external source into FPF-native inputs.Shape (conceptual):
⟨ MapperId, ExternalIndexId, MappingPolicyRef, Targets{ClaimSheet|BridgeHints|SoSFeatureSet|UTSProposals}, PlaneMapRef?, ScaleEmbeddingSpecRef?, EvidenceGraphId?, CSLCProofStubs? ⟩Notes.
- The mapper cites the governing definitions (
A.19,G.0,G.3,G.4) and publishes the pins needed for audit and refresh. It supplies no separate legality gate. - When cross-plane or cross-context reuse is implicated, alignment outputs must cite the actual relation and the receiving-use and reliance grounds required by its governor. Carry bundle or gate anchors when an independently applicable
E.18flow crossing orA.21gate consumes the alignment, as required byCC‑GCORE‑CROSS‑1. - Avoid “edition echo”: downstream artefacts cite
ExternalIndexRef.editionandClaimMapperRef.edition(and optionalPlaneMapRef.edition/ScaleEmbeddingSpecRef.edition) rather than copying snapshot ids/editions as free fields.
- The mapper cites the governing definitions (
-
SoSFeatureTransform@Context— declares how external signals become CHR‑typed SoS features (for DHC/dashboard usage and/or SoS‑LOG rule evaluation).Shape (conceptual):
⟨ SoSFeatureTransformId, Inputs{ClaimSheetId[] | ExternalSignalsRef}, SoSFeatureSetId, FeatureTypingRefs{CharacteristicId/ScaleId/CoordinateId}, ReferencePlane, EvidenceGraphId?, PathSliceId[]?, ProofHooks? ⟩Notes.
- The derivation records typing + provenance; comparator and legality definitions remain with the cited governing patterns. When features are used as DHC measurements, establish or reuse C.16 measurement results. Make the active C.21
DHCReplayBasisrecoverable for every persisted, compared, aggregated, or published coordinate.
- The derivation records typing + provenance; comparator and legality definitions remain with the cited governing patterns. When features are used as DHC measurements, establish or reuse C.16 measurement results. Make the active C.21
-
ScaleEmbeddingSpec@Context— optional constraints for representation/space alignment used inside an alignment recipe.Shape (conceptual):
⟨ ScaleEmbeddingSpecId, IntendedUse, AllowedTransformFamily, RequiredPins{NormalizationMethodRef.edition?}, ProhibitedCoercions ⟩Design intent. State the constraints and pin the edition for each representation alignment. If the transformation introduces a new scale, establish its lawfulness under the applicable scale rules.
LEX/UTS note (informative).
ScaleEmbeddingSpecis a LEX head. When you mint a public id for one, publish its UTS row under F.17 with the Tech designation and any Plain designation needed for the stated didactic use. -
IndexTelemetryPin— an emitted refresh input that makes interop changes RSCR‑visible.Shape (conceptual; RSCR‑typed):
⟨ triggerKindId: RSCRTriggerKindId, scope: PathSliceId[] | PathId[] | PatternScopeId, payloadPins{ExternalIndexId, ExternalIndexRef.edition, ClaimMapperRef.edition?, MappingPolicyRef?, PlaneMapRef.edition?, ScaleEmbeddingSpecRef.edition?, PathId[]?, PathSliceId[]?, UTSRowId[]?, …} ⟩Publication. Emitted to
G.11as refresh input; recorded with canonicalRSCRTriggerKindIdcauses. -
InteropSurface@Context— a selector-facing or dashboard-facing summary of what interop publications and records exist and how they are pinned.Shape (conceptual):
⟨ InteropSurfaceId, ExternalIndexId, ExternalIndexRef.edition, MapperId?, ClaimMapperRef.edition?, MappingPolicyRef?, SoSFeatureSetId?, EvidenceGraphId?, PathSliceId[]?, PlaneMapRef.edition?, ScaleEmbeddingSpecRef.edition?, UTSRowId[] ⟩Publication. Published to UTS with twin labels as applicable.
G.13:4.3 - Use external sources to construct the needed input
Start with the input required by the receiving method and records from the external source. Reuse an adequate mapping recipe. To construct or revise one, establish the transformation and its limits before publishing its card and routing its outputs.
G.13:4.3.1 - Construct and check the mapping
- Recover the required output meaning. Read the receiving operation’s input definition and conditions. For a G.2 ClaimSheet, this includes the claim, its subject, scope and source support; for a CHR feature, the subject, characteristic, scale, units and observation basis come from G.3/C.16. For a Bridge hint, preserve the two local meanings and the proposed relation for F.9 to examine. Select only the outputs needed by the receiving question.
- Recover what the available source fields denote. Inspect actual records together with the provider’s definitions for that edition: the counted or described object, population, inclusion rules, units, dates/windows, missing-value conventions and identifiers. Follow a record’s citation to the underlying source when its metadata does not contain the claim or conditions needed. Use the name to locate the field. Its definition and examples establish what a value represents. If the definition is unavailable, ask the provider or obtain the specified source passage before using the ambiguous field for that output.
- Connect each required output to its basis. Select the source passage or field that supplies the output, or define the operation that derives it. Direct copying requires compatible meanings. A count requires the population, inclusion predicate and identity/equivalence rule; a unit conversion requires a lawful scale conversion; a claim extraction retains the source’s subject, conditions and contrary evidence under G.2. For a semantic correspondence use F.9’s relation construction and its separate receiving-use test. For a measurement use C.16’s measurement model and G.3’s lawful operations. If those methods and the available domain methods leave a transformation unspecified, identify what is missing and the source or competent supplier needed to construct it.
- Apply the recipe to real records. Work through at least one representative record to the required result and inspect a consequential variation, such as a duplicate identifier, an absent value, a changed time window or a different population definition. Choose the variation from the assumptions the recipe uses. Retain the source value and derivation sufficiently for the receiving use to check them. Missing or ambiguous values remain missing or ambiguous; they become zero, false or a negative claim only when the source meaning and derivation establish that result.
- Return the input or identify what prevents obtaining it. A mapped result carries its source edition, applicable scope and transformation basis. A failed compatibility condition supports rejection of that mapping. An unresolved field definition, relation, measurement procedure or source claim prevents only the outputs that depend on it. Identify what source, observation or method construction could settle the gap and what can already be used. When the receiving contract requires a complete bundle, retain that unmet condition rather than treating a partial mapping as the bundle.
Worked mapping. Suppose a review needs the number of distinct research articles in a named collection published during 2025, and source-grounded ClaimSheets about their methods. The available index supplies records with work_id, year, record_type, a citation and an abstract. In this illustrative provider contract, year means first publication year, work_id identifies a work across duplicate records, and record_type = research_article is the relevant inclusion class.
The recipe filters the stated collection by year and class, groups qualifying records by work_id and counts those groups. Three qualifying records with ids w1, w1, w2 establish two works in that slice. If both belong to a larger collection, at least two works in it qualify. Its exact total needs complete coverage or independent evidence accounting for the remainder. For example, independent evidence that exactly one further work qualifies outside the slice establishes three in all. Finding one additional work while the rest are unknown establishes a minimum of three. Use as a DHC measurement additionally establishes the C.16 result and C.21 replay basis.
For the ClaimSheets, the citation leads to the article text. The reviewer extracts the method claim with its tested population and conditions under G.2; an abstract that omits those conditions leaves an extraction gap to settle from the article. The count and metadata alone supply no performance claim.
If the provider instead defines work_id as a record identifier, the same three values no longer establish two distinct works: obtain an identity rule or source-level matching method before returning the work count.
If year means index ingestion year, the recipe must obtain publication dates to answer the original question. Changing the receiving question to the number of index records ingested in 2025 can make those fields sufficient for that different result.
G.13:4.3.2 - Publish and use the established recipe
-
Register source editions. Author
ExternalIndexCard@Contextfor each external source snapshot mapped through this kit, includingExternalEditionand theentityOfConcernplane anchor. -
Author mapping recipes. Create
ClaimMapperCard@Contextfor the established transformation and its limits, identifying the produced artefacts and applicable policy/spec references. IncludePlaneMapReforScaleEmbeddingSpecRefwhen the transformation uses them. -
Produce FPF-native inputs. Apply the recipe and route its obtained outputs to the receiving operation, for example:
G.2harvesting (ClaimSheets / operator & object inventories / candidate bridge hints),G.3CHR typing (when numeric signals are formalized as CHR characteristics/scales/coordinates),G.4acceptance/threshold policies (when a downstream decision requires explicit CAL policy rather than telemetry),G.12dashboards (when derived SoS features support DHC readings; measurement claims consume C.16 results with the C.21 replay basis required by their use).
-
Use the resulting inputs under the receiving rules. Open only the applicable branch:
- When selection or a selector-facing declaration is needed,
G.5consumes the applicable grounded inputs and returns its declared outcome: a set result, narrowed handoff, abstain or escalation. - When a reproducible parity report is needed,
G.9consumes the pinned editions and windows under its comparison plan. - When shipping is needed,
G.10may include interop surfaces as cited publications or records.
- When selection or a selector-facing declaration is needed,
-
Emit telemetry and refresh causes. On any change in external editions, alignment policies, plane maps, or embedding specs, emit:
- a canonical
RSCRTriggerKindId(perG.Core), - the affected scope (
PathSliceId[],PathId[]orPatternScopeId), - payload pins (editions, policies and UTS rows).
With these pins,
G.11can identify dependent uses and decide whether their changed basis calls for further work. - a canonical
G.13:4.4 - Interfaces — minimal I/O standard (conceptual; kit‑only)
| ID | Interface | Consumes | Produces |
|---|---|---|---|
G.13‑1 Register_ExternalIndex | Register ExternalIndexCard@Context | Provider metadata, scope, ExternalEdition, freshness, entityOfConcern anchor | ExternalIndexCard@Context (+ UTS row when published) |
G.13‑2 Map_ClaimsToFPF | Apply ClaimMapperCard@Context | ExternalIndexCard@Context, MappingPolicyRef, optional PlaneMapRef/ScaleEmbeddingSpecRef, optional EvidenceGraph hooks | ClaimSheet@Context, BridgeHints, optional SoSFeatureSet@Context, optional UTS proposals |
G.13‑3 Derive_SoSFeatures | Produce CHR‑typed SoS features | ClaimSheets / external signals refs, CHR typing refs, legality proof hooks | SoSFeatureSet@Context (CHR‑typed; provenance pinned) |
G.13‑4 Publish_InteropSurface | Publish interop summary | outputs of G.13‑2/‑3, UTS refs | InteropSurface@Context (+ UTS rows/twins) |
G.13‑5 Emit_IndexTelemetryPin | Emit refresh input | edition/policy changes + scope + payload pins | telemetry to G.11 (typed causes + payload pins) |
G.13‑6 Wire_To_SoTA_Pack | Provide shipping hook | InteropSurface@Context + citations to upstream artefacts | G.10 pack hooks (as cited payload; no serialisation mandated) |
G.13:5 - Extensions (pattern‑scoped; non‑core)
Use GPatternExtension blocks with stable PatternScopeIds for provider- and method-specific requirements. The blocks identify the required pins, editions and policies. Cite the governing pattern and the provider or method that supplies the meaning or transformation used.
G.13:5.1 - G.13:Ext.ExternalIndexProviderWiring
- PatternScopeId:
G.13:Ext.ExternalIndexProviderWiring - GPatternExtensionId:
ExternalIndexProviderWiring - GPatternExtensionKind:
InteropSpecific - GoverningPatternId:
G.13(source registration and refresh wiring) - Uses:
{G.13} - ⊑/⊑⁺:
∅
RequiredPins/EditionPins/PolicyPins (minimum):
ExternalIndexTypeExternalEdition(as published onExternalIndexCard@Context)Licence?CoverageScopeProviderChangePolicyId?(if provider‑specific “schema drift” handling exists)
RSCRTriggerSetIds / RSCRTriggerKindIds: ∅ (covered by G.13:4.1)
Notes (wiring-only):
- Provider‑specific ingestion choices (e.g., OpenAlex‑class, Crossref‑class, ORKG‑class, discipline repositories) must not become Part-G-wide norms. This module only records which provider cards exist and which editions/policies are pinned.
G.13:5.2 - G.13:Ext.EmbeddingBasedAlignment
- PatternScopeId:
G.13:Ext.EmbeddingBasedAlignment - GPatternExtensionId:
EmbeddingBasedAlignment - GPatternExtensionKind:
MethodSpecific - GoverningPatternId:
G.13(mapping interface); the selected alignment method governs its transformation and qualification. - Uses:
{G.13, A.19, E.5.2} - ⊑/⊑⁺:
∅
RequiredPins/EditionPins/PolicyPins (minimum):
ScaleEmbeddingSpecRef.editionNormalizationMethodRef.edition?(when a declared normalization/representation transform is used)MappingPolicyRefEvidenceGraphId?(when evidence paths for alignment decisions are published)
RSCRTriggerSetIds / RSCRTriggerKindIds: ∅ (covered by G.13:4.1)
Notes (wiring-only):
- A selected embedding-based method supplies the transformation and its qualification for the receiving relation. Cite that method and its edition with the applicable
ScaleEmbeddingSpecandNormalizationMethodreferences.
G.13:5.3 - G.13:Ext.EntityResolutionAndAliasDocking
- PatternScopeId:
G.13:Ext.EntityResolutionAndAliasDocking - GPatternExtensionId:
EntityResolutionAndAliasDocking - GPatternExtensionKind:
InteropSpecific - GoverningPatternId:
G.13(mapping interface),F.17(public naming); the selected entity-resolution method governs the identity judgement. - Uses:
{F.17, E.10} - ⊑/⊑⁺:
∅
RequiredPins/EditionPins/PolicyPins (minimum):
UTSRowId[](for externally‑sourced entities that become publicly citable)ExternalIdAliasSetId?(labels only; canonical ids remain UTS ids)TokenizationPolicyId?
RSCRTriggerSetIds / RSCRTriggerKindIds: ∅ (covered by G.13:4.1)
Notes (wiring-only):
- Establish source-entity identity through the selected resolution method before docking its external identifiers as aliases. Preserve the basis for identifying two records as one entity, or leave that identity unresolved; publishing an alias does not establish it.
G.13:6 - Archetypal grounding
System. Software architecture portfolio design.
Register an external scholarly index edition for “software architecture” concept neighborhoods. Align extracted technique/tactic claims into ClaimSheets and derive a CHR‑typed feature set (e.g., evidence depth, maturity). Select a set of tactics under multi‑objective tradeoffs, use G.5 to declare that result, and ship a SoTA pack that cites the interop surface.
Episteme. Science‑of‑science discipline dashboard.
Align external claim graphs (replication, standardisation, disruption-style proxies) into CHR-typed features. For the selected DHC coordinates, establish or reuse C.16 measurement results and recover their active C.21 DHCReplayBasis. Use those result refs in the DHC series and its dashboard slice. Publish the selected series episteme through the dashboard form, citing ExternalIndexRef.edition, ClaimMapperRef.edition, and MappingPolicyRef; emit refresh triggers when the external edition updates.
OEE/QD. Open‑ended environment generation. Register external environment/task taxonomies as index cards. Align them into generator‑family registries (as cited publications or records), using coverage and regret only as telemetry in this example. When the taxonomy edition changes, use G.11 to assess the mappings and downstream uses that depend on it; realign those whose basis no longer suffices.
G.13:7 - Bias‑Annotation (informative)
- Vendor/tool bias. The kit names conceptual surfaces only; it avoids vendor‑specific file formats or tooling claims.
- Metric‑authority bias. External indicators are treated as inputs that must be typed, pinned, and evidenced; they are not authority by default.
- Representation bias. State which representation or embedding specification was used and pin its edition so the reader can recover the meaning of the alignment.
- Discipline bias. Interop supports pluralism by preserving explicit crossings and versioned alignments instead of forcing a single canonical external ontology.
G.13:8 - Conformance Checklist (CC‑G13; applies when G.13 surfaces are used)
-
CC‑G13‑CoreRef. (normative)
G.13implementations MUST satisfy the effectiveG.Coreobligations declared byG.13:4.1(GCoreLinkageManifest), including trigger typing, Default Governing Definition Index citation, and crossing‑visibility pin discipline. -
CC‑G13‑InteropIsNotASpecRefSurface. (delegated) Interop surfaces cite the CN‑Spec/CG‑Spec/CHR/CAL definitions and pins applicable to their transformation and receiving use. They MUST NOT introduce separate legality or comparability gates. → delegate to
CC‑GCORE‑CN‑CG‑1. -
CC‑G13‑CrossingsAreExplicitWhenInteropTouchesPlanesOrContexts. (delegated) For reuse across planes or contexts, the alignment MUST identify the obtaining relation and the receiving-use and reliance grounds required by its governor. An independently applicable E.18 flow crossing or A.21 gate MUST retain its required visibility anchors. → delegate to
CC‑GCORE‑CROSS‑1. -
CC-G13-PlanePenaltyPoliciesArePinned. (local; governing-definition citing) When a receiving use applies a numerical plane loss, interop surfaces MUST publish the relevant model and policy-id pins under the crossing-visibility discipline, and those policies MUST satisfy the constraints governed by
CG-Spec(citeCC-G0-Φ). APlaneMapRefalone supplies no loss model or penalty. Interop surfaces MUST NOT define interop-local penalty functions. -
CC‑G13‑SetReturnPreserved. (delegated) Interop MUST NOT introduce hidden scalarisation or forced single‑winner selection. → delegate to
CC‑GCORE‑SET‑1. -
CC‑G13‑DefaultClaimsAreCitationsOnly. (delegated) Any mention of defaults (e.g., dominance regime,
PortfolioMode) is a citation to the default’s governing definition throughG.Core.DefaultGoverningDefinitionIndex, not a local default statement. → delegate toCC‑GCORE‑DEF‑1. -
CC‑G13‑EditionDisciplineForInteropCards. (local)
ExternalIndexCard@ContextandClaimMapperCard@ContextMUST expose edition pins (ExternalIndexRef.edition,ClaimMapperRef.edition). Any interop surface published to UTS MUST cite the relevant…Ref.editionvalues (incl.PlaneMapRef.edition?,ScaleEmbeddingSpecRef.edition?) when present. FPF edition keys MUST appear only on…Ref.editionpins when a reference is present. Provider snapshot labels (e.g.,ExternalEditiononExternalIndexCard@Context) may exist on the source card, but MUST NOT be copied into downstream artefacts as free‑floating “edition fields”; downstream artefacts cite the corresponding…Ref.editionpins instead. -
CC‑G13‑SoSFeaturesAreCHRTypedAndLegal. (local; governing-definition citing) If
SoSFeatureTransform@Contextis used, produced SoS features MUST be CHR‑typed viaFeatureTypingRefs{CharacteristicId/ScaleId/CoordinateId}(governed byG.3) and any legality/units obligations must be satisfied via CSLC/CG governing definitions (citeA.18/G.0/G.4; do not invent interop‑local legality gates). When those features serve as DHC measurements, the use MUST consume C.16 measurement results; every persisted, compared, aggregated, or published coordinate MUST make its active C.21DHCReplayBasisrecoverable. -
CC‑G13‑TelemetryEmitsCanonicalTriggerKinds. (delegated) Interop‑driven changes (external edition bumps, mapping policy changes, plane‑map edits, embedding‑spec edits) MUST emit canonical
RSCRTriggerKindIdcauses with scope and payload pins. → delegate toCC‑GCORE‑TRIG‑1,CC‑GCORE‑TRIG‑2,CC‑GCORE‑TRIG‑3,CC‑GCORE‑TRIG‑4. -
CC‑G13‑IDContinuityForExternallySourcedIdentifiers. (delegated) Interop publication MUST follow Δ‑discipline: no “renaming by meaning”; use aliases/deprecations as required. → delegate to
CC‑GCORE‑ID‑1,CC‑GCORE‑ID‑2. -
CC‑G13‑NotationIndependence. (local) Conformance is judged on the conceptual objects in
G.13:4.2. Any serialisation is non‑normative and must not redefine semantics. (CitesE.5.2for notation independence.)
Interop transforms MUST NOT perform illicit arithmetic on ordinal/compare‑only scales (e.g., averaging or subtraction); any aggregation must be via lawful CAL operators with explicit scale legality (cite A.18 / CC‑G0‑CSLC).
G.13:9 - Common Anti‑Patterns and How to Avoid Them
-
Anti‑pattern: “Format == spec”. Treating an export schema (KG dump, JSON, RO‑Crate, etc.) as the normative definition. Remedy: Keep
ExternalIndexCard/ClaimMapperCard/InteropSurfaceas the conceptual specification; treat serialisation as an appendix/tooling concern. -
Anti‑pattern: Hidden scale invention. An embedding similarity becomes a “score” without explicit typing/binding. Remedy: Require
ScaleEmbeddingSpecRef+ edition pins and bind any derived features through CHR/CAL governing definitions. -
Anti‑pattern: Implicit plane/context reuse. Using an external concept mapping without its actual relation and receiving-use grounds. Remedy: Establish the actual sense, kind or plane relation under its governing rule and retain the receiving-use grounds. Publish the bundle or gate anchors required by any actual E.18/A.21 use.
-
Anti‑pattern: Edition‑free dashboards. Feeding externally derived rows into dashboards without pinned editions/policies. Remedy: Pin
ExternalIndexRef.editionandClaimMapperRef.edition; emit RSCR triggers on changes. -
Anti‑pattern: Interop asserts defaults. “Interop decides dominance regime /
PortfolioMode.” Remedy: Treat defaults as citations only (the relevant governing definition is cited throughG.Core.DefaultGoverningDefinitionIndex).
G.13:10 - Consequences
- Changes become locatable. External source changes produce typed RSCR causes with scope and payload pins for reconsidering dependent uses.
- External material supports construction. External claims and features can support SoTA synthesis, candidate exploration and audit with their source basis recoverable.
- Recoverable source meaning. Source and mapper cards, with their edition pins, help keep a provider, tool or format from changing how the result is interpreted.
- Cross‑tradition reuse becomes auditable. The alignment exposes the relation used across planes or contexts and its receiving-use grounds. An independently applicable flow crossing or gate retains its required visibility anchors.
G.13:11 - Rationale
An interop kit lets authors use external scholarly sources while keeping comparisons, editions and transformations visible.
G.13 connects source meanings and records to receiving inputs through :4.3, then makes the transformation and source editions citable. Its cards and surfaces retain the applicable provenance and refresh inputs. Extensions bind provider or method details used by that transformation.
G.13:12 - SoTA-Echoing (informative)
-
Scholarly claim graphs & open indexes. Research knowledge graphs and scholarly indexes can supply claim representations and concept taxonomies as mapping inputs. Treat these as sources registered via
ExternalIndexCard, not as governing patterns. -
Neural representations for scientific text. Scientific encoders, such as SciBERT-class and citation-aware SPECTER-class representations, are candidates for alignment heuristics whose suitability must be established for the receiving question. In FPF terms, they belong behind
ScaleEmbeddingSpec+ pinned editions/policies (seeG.13:Ext.EmbeddingBasedAlignment). -
Schema matching and entity resolution. Candidate approaches include deep entity matching, contrastive representation alignment and graph alignment. Select a mapping procedure for the stated relation and evidence, then record its use and limitations through the applicable extension.
-
Systematic-review reporting. PRISMA-style records make source selection and exclusions inspectable. They can supply provenance and coverage inputs; G.2 still judges coverage for the receiving question. Preserve source editions and windows when consuming those records.
-
QD / Illumination and OEE declared set results. When a QD or OEE use consumes an external taxonomy or environment corpus, expose its edition and mapping basis. Coverage and regret remain telemetry unless the applicable governing rule explicitly promotes them into the decision criteria.
G.13:13 - Relations
Builds on: G.Core.
Imports: G.2, G.3, G.4, G.5, G.6, G.7, G.9, G.10, G.11, A.19, A.18, G.0, F.17, E.5.2, E.18.
Publishes to: UTS (twin labels where applicable); refresh inputs to G.11; shipping hook surfaces to G.10 (as cited publications or records).
Relates to: G.12 (dashboards), G.8 (SoS-LOG bundle surfaces) when interop-derived publications or records are consumed there.
G.13:14 - Author’s quick checklist (informative)
- Establish the needed input and mapping from source definitions and actual records under :4.3; register each source snapshot with its
ExternalEdition. - Author a
ClaimMapperCard@Contextwith theMappingPolicyRefand required edition pins. - If you derive SoS features, declare a
SoSFeatureTransform@Contextand cite CHR typing refs and provenance hooks. - Publish an
InteropSurface@Contextthat cites all active…Ref.editionvalues and UTS rows. - On any external edition or policy change, emit canonical RSCR trigger causes with the affected scope and payload pins.
- Keep provider and method conditions with the applicable extension and its cited procedure.