CC-E23-1 | Name the exact object version, exact object-under-improvement evaluation, one current QualityEvaluationQuestionFrame, and one QualityEvaluationUseDeclaration before claiming a changed evaluation result. |
CC-E23-2 | Reuse an E.22 or equivalent exact frame only when it binds the current object version, selected characteristic space, predicate or comparator, ClaimScope, result-consuming work or decision, purpose, qualification window, and any grounded non-use boundary required by E.22; otherwise open a new frame. |
CC-E23-3 | Represent returned repair possibilities as row-atomic E.22 findings or proposal rows with closure tests; pair proposals selected for the next pass with scale-qualified predicted evaluation-result changes, using ExpectedEvaluationResultChange@Context when the conditional §4.2a account is needed. A grouped memory summary does not discharge skipped rows, and proposal selection does not establish performance. |
CC-E23-4 | Every asserted evaluation or improvement U.Work first recovers each exact actual performer through A.13, then uses A.15.1 to identify the occurrence, time, Method, and containing System independently. Add A.2.1 and F.6 only when the record or receiving use expressly represents precise assignment-bound attribution; their absence or failure leaves the Work intact. Then name the evaluation application and result binding or direct result or change relation, plus any separate result episteme. Re-evaluate the changed object before claiming coordinate, status, Q, or front-relation change. |
CC-E23-5 | Record what became worse and protected trade-offs. |
CC-E23-6 | Continue only when a scale-qualified expected evaluation-result change and the cost and risk account support another pass. |
CC-E23-7 | Decide a local stop from the achieved aim, protected trade-offs and absence of a worthwhile feasible improvement under §4.3. All-5, exceptional or front-reaching results neither require nor suffice for that decision. |
CC-E23-7a | Do not treat 5, all-5, or 5-defensible as a repair target. Repair below-floor results first. Exceptional-improvement work proceeds through non-dominated proposal rows that name the expected substantive content change, protected trade-offs, and cost and risk. A no-proposal or stay-at-current-value disposition is admitted only when it names the checked evaluation evidence basis, using LoopEvaluationEvidenceBasis@Context when the conditional structured account is needed, and explains why every plausible content improvement is dominated, unavailable, or outside the declared scope. Reject changes that add guards, relation catalogues, evidence theatre, or quality proof while reducing use, affordability, locality, or ecology. |
CC-E23-8 | When a neighboring claim appears during a loop, name the live claim and its subject pattern before continuing. E.23 may cite that pattern in the loop record, but it does not absorb the neighbor’s authority unless the neighbor’s object version is itself the declared object under improvement. |
CC-E23-8a | For a precision-restoration defect, apply F.19 as guidance and open a subject pattern only for unresolved FPF-specific meaning; claim Method or MethodDescription only after A.3.1 and A.3.2 admit it. Apply CC-E23-4 only when actual repair Work is asserted. Consume E.21’s compact PrecisionRestorationProfile when that evaluation is active. Require one bounded KindRestorationCheck when the changed expression can alter the object, kind, relation, slot or use position, claim kind, admissible use, or scope; otherwise F.19’s local revalidation completes the ordinary repair. |
CC-E23-9 | Apply E.10 to load-bearing loop names, status values, examples, stop conditions, and result wording introduced or repaired by the loop. |
CC-E23-10 | Preserve the named evaluation’s evidence basis, result-row shape, short-rationale rule, required result summaries, and coordinate-specific payloads in every re-evaluation. For E.21, consume the compact PrecisionRestorationProfile under E.21:4.3a. |
CC-E23-11 | If a practitioner entry phrase such as “loop engineering”, “agent loop”, or “harness loop” appears, lower it to object version plus object-under-improvement evaluation before opening E.23, or name the direct neighboring subject pattern and stop the E.23 overread. |
CC-E23-12 | In agent or harness cases, state which slice the loop may change: the target object version, the evaluation, or the harness object. Any other slice becomes neighboring work under its own subject pattern, not implicit E.23 scope. |
CC-E23-13 | Keep the selected proposal, actual improvement Work governed by CC-E23-4, its result or change relation, changed object or Transformation, later evaluation pass, and result episteme distinct. When a required relation has no governor, retain those objects and the blocker; do not mint a generic Work-result relation. |
CC-E23-14 | When §4.2a’s receiving-use condition holds, represent current alternatives, exact guards, selected obtaining relations, selected continuation, stop, and subject-assertion reconsideration conditions in one admitted A.22 unfolding structure. When transformation-flow membership is current, E.18/E.18.3 recognizes that same selected U.Structure; do not mint a parallel loop object. A visible cycle, record, structure, decision value, or branch is not enduring Work, context, authorization, or performance. |
CC-E23-15 | A low value, finding, floor miss, or improvement aim does not establish an actual Problem. Any actual Problem used by the loop resolves to one current C.22.PFR occurrence with its direct participants and temporal identity. |