A.15.1:11 - Common Anti-Patterns and How to Avoid Them

“The log is the performed occurrence.” Dumping telemetry without recoverable candidate-action facts—every actual performer System’s A.13 basis, at least one Method actually followed, time window, and at least one obtaining local containing-system relation, plus any other relation the receiving claim uses—does not establish Work. Recover and admit the occurrence independently, keep the log as evidence, and add F.6 only afterward when precise assignment-bound attribution is current.
Record-handling-as-transformation. ETL, copying, formatting, evaluation, or publication work is treated as proof that a record or dataset changed -> Keep the grounded Work occurrence, but assert actual change only after A.3.4 identifies the transformation and either a declared domain predicate with the exact Work and transformation participants obtains or a C.2.1 local compound claim under A.6.RCD disposition 2 exposes its substrate-admitted constructor, governed base predicates, actual participants, and case facts; otherwise return missing-governor[work-to-change].
Silent cross-locality acceptance. “Ops accepted it, so audit accepts it.” -> Name each receiving criterion, evaluation work, and result episteme. Assert acceptance only through that use’s declared predicate and actual participants; otherwise return missing-governor[acceptance]. If the criteria use different local senses, test the F.9 Bridge, state the proposed cross-local comparison or substitution in a separate bounded-use claim, and check reliance; the Bridge itself transfers no acceptance.
Description-change-as-occurrence-change. Selecting another MethodDescription episteme is treated as automatically splitting or preserving Work -> State the description-selection change separately. Only when an accompanying actual history change creates an identity question for a named use should its continuity-policy criterion be applied; the policy revises the judgment, not the occurrence. Call the descriptions editions only when their exact C.2.1 relation obtains.
Budget on the method or system-role object. Charging costs to a Method, local system-role kind, or assignment -> Attribute performed resource use only through exact relations involving Work individuals; keep estimates in Method descriptions or plans.
Part ambiguity. Mixing retries, episodes, and operational parts with no declared relation → Choose and declare the part relation.
Timetable-as-Work-architecture. Rows with similar labels or one planned window are treated as one Work whole, its parts, or overlapping Work → Recover every actual Work occurrence first; then establish each Work-part and temporal relation separately. Keep an unperformed or ungrounded row as plan or description content.
Slice-as-episode. A monitoring interval, telemetry window, crank-angle segment, or one-second reception trace is called an episode only because it has timestamps -> Keep it as a C.27.TA temporal aspect, evidence relation, or telemetry relation. Use TemporalPartOf_work or EpisodeOf_work only after the first participant is independently admitted as Work and the corresponding §4.1a predicate passes; add a continuity policy only if direct boundary facts leave its grouping ambiguous.
Episode-as-new-work by habit. A pause, retune, or interruption is always recorded as either a new occurrence or the same one -> Preserve the boundary events first. Apply exact workContinuityPolicyRef only when a named use must decide the grouping; otherwise return unresolved segmentation rather than forcing either answer.
Method-factor-as-work-part by label. A step, stroke, receiver component, graph node, or method-description section is treated as a work part or submethod by name -> Recover the current object: U.Method factor, U.MethodDescription constituent, TemporalPartOf_work, OperationalPartOf_work, evidence segment, mechanism material, system-component behavior, or missing-source-relation note.
Granularity inflation. Every interval or trace row receives a durable work-part name -> Name the work part only when a current resource, evidence, KPI, acceptance, repair, aggregation, cross-context reliance, or source-relation return use hangs on it.
Union-hull confusion. Changing KPI coverage silently between reports -> recover the exact B.1.4 temporal aggregation and cite its policy per KPI.
Double-count in overlaps. Summing child and parent resource facts as one ledger -> recover the B.1.6 aggregation claim and apply its exact overlap or deduplication policy.