feat(analysis): bind inferred-status refusals to an analysis-run profile - #473
Conversation
GAP-004 leftover / ADR 0073. Bind existing EvidenceStatus, refuse_inferred_as_observed, and refuse_inferred_as_transition to cutoff-safe inferred_status_v1. Mixed observed/inferred corpora emit a digest-bound census. identity_recovery_rate stays library-side. Fixes #472.
seonghobae
left a comment
seonghobae
left a comment
There was a problem hiding this comment.
Operator-visible GAP-004 leftover / ADR 0073. Binds existing inferred_status::EvidenceStatus + refuse_inferred_as_observed + refuse_inferred_as_transition to inferred_status_v1 (tepp.inferred_status.v1). Inference inferred_is_not_observed_and_not_transition. AvailableTime cutoff exclusion and MAX_EVIDENCE_UNITS enforced. identity_recovery_rate stays library-side. unobserved and no_relationship are not wire statuses. Distinct from #461 episode-membership, #460 relation-absence, #458 outcome-order, #434 membership-target, #430 location-membership, and #416 copy-identity. Author COMMENT is not independent APPROVE. Two independent current-head APPROVEs required. Exact-head Checks on 3c66702 only. Do not self-approve.
📝 WalkthroughWalkthrough
Changes추론 상태 분석 실행 프로필
Estimated code review effort: 4 (Complex) | ~45 minutes Merge Risk: 🟡 Moderate · up to This PR adds a public inferred-status analysis profile and artifact, but the current implementation can accept an artifact claiming more evidence than the enforced limit, while authorization/provenance and failed-run completion remain dependent on callers. It is not merge-ready until the limit validation is fixed and the boundary and failure-handling ownership are explicitly addressed. Sequence Diagram(s)sequenceDiagram
participant 실행요청
participant execute_inferred_status_run
participant inferred_status
participant 종료결과
실행요청->>execute_inferred_status_run: 요청·수락·snapshot·cutoff 검증
execute_inferred_status_run->>inferred_status: 증거 집계 및 inferred 상태 거부
inferred_status-->>execute_inferred_status_run: 관찰·추론 census와 거부 카운터
execute_inferred_status_run->>종료결과: SHA-256 digest-bound 결과 생성
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
Full details: Linked Issues checkExplanation PR은 Full details: Docstring CoverageExplanation Docstring coverage is 51.72% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 29 functions across 3 files. (7 skipped: 7 unsupported.)
✨ Finishing Touches 💡 1📝 Generate docstrings 💡
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![devin-ai-integration[bot]](https://avatars.githubusercontent.com/in/811515?s=60&v=4){kind=small}
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
Use the review-reproduced oversized compact artifact as RED evidence. Enforce the same MAX_EVIDENCE_UNITS invariant at artifact deserialization/serialization that execution already enforces, and add regression coverage for both directions.
Document the artifact-count invariant and exact-head landing procedure, and collapse the duplicated documentation map into one canonical index so the new doctoring entry cannot diverge between copies.
seonghobae
added
enhancement
ChatGPT Codex Connector
e0e4480
into
feat/copy-identity-analysis-run-gap-004
1 participant
Consolidation status
fold_into_landing_vehicle— this PR is now retargeted directly onto the provisional Validation / Analysis Run simple-refusal landing vehicle #416 (feat/copy-identity-analysis-run-gap-004). It binds the existinginferred_statusrefusal family into that bounded-context vehicle; it is not an independently shippable bounded context. Preserve its source, focused tests, doctoring and review/RED-GREEN evidence before closing it. ADR 0073 remains implementation lineage pending #437 normalization, not per-rule architecture authority.Current micro-profile exact head:
9ae0488e44bc0775b2289c71d83d88c57e660b0d.Current landing base at retarget:
#416@a6ba5d79c8d9acc1e7b4a53b3be5ef3d70ab19e7.GitHub currently reports this retargeted stack as non-mergeable because #416 and this branch independently changed the shared Analysis Run integration files (
Cargo.toml,lib.rs, lockfile and common documentation/index surfaces). That is a consolidation conflict, not permission to merge the micro-profile independently. Resolve it by preserving #416's copy-identity repairs and this branch's inferred-status implementation/tests in one landing head; do not overwrite either side and do not close this PR until the surviving head contains the unique source/test evidence.Current repair
Fresh review found that execution enforced
MAX_EVIDENCE_UNITSbut a compact serializedInferredStatusArtifactcould claim a larger census because artifact validation lacked the same upper bound. The repair on this branch makes serialization/deserialization enforceevidence_count <= MAX_EVIDENCE_UNITS, with a compact max+1 regression proving an impossible census fails closed without allocating an oversized evidence vector. The documentation repair also removed the duplicated secondDOCUMENTATION.mdmap and made the exact-head hosted verification rule code-current.The profile preserves:
AvailableTimeexclusion;observed/inferredcensus;identity_recovery_ratestays library-side andscientific_acceptanceis not exposed;Landing gate
After conflict resolution, the surviving #416 exact head must reacquire hosted Rust/documentation/security/SAST evidence and qualifying independent review. Queued, stale, predecessor-head, or this micro-profile's prior evidence does not transfer. Do not self-approve, bypass protection, force-push, or treat this Draft as a production dependency.