completion-audit

Completion Audit

When to use

Wrapping up a long / multi-stage task, or any task that produces user-visible artifacts (runs, dashboards, reports, exports).

Enumerate the touched files / user-visible surfaces this task changed.
Reconcile: compare the plan/objective against expectations expressed in the thread (not just what's written in the plan). Differences are surfaced, never silently dropped.
Map each touched surface to a check in the protected acceptance_manifest.yaml (read-only). Any user-visible surface with no check → record in review_queue.yaml and mark blocked.
Gather real evidence paths (actual artifacts/configs), not run_state assertions.
Write completion_candidate.yaml (your proposal; status = candidate_complete).
Invoke the external gate: check_acceptance.py --manifest ... --inventory ... --candidate ... --repo .

completion_candidate.yaml (proposal only)
review_queue.yaml (blocking open items)
gate result (PASS → an external verifier may grant complete; FAIL/any review item → blocked)

Do not set complete yourself — you may only reach candidate_complete or blocked. Only the external gate/verifier writes complete.
Do not reclassify a real regression as "subjective" to skip it — needs-review == blocked.
Do not feed the gate your run_state as evidence — the gate must read real artifacts.