Give investigation concerns a disposition

What the check reads

The default-on verify-stage check emits warnings for open concerns whose review_due date is due or whose review_due field is absent. A missing concerns directory passes because there are no records to inspect. Promoted and waived concerns are no longer open for this check. Supported review dates use YYYY-MM-DD or RFC3339. Reported due concerns are actionable findings rather than advisory-only diagnostics.

proof audit --check concern_adjudicated

An illustrative case

Illustrative case: a reviewer suspects a retry path may submit the same transaction twice, but has not reproduced it. Record the suspected condition, relevant source and an investigation date. When the date arrives, investigate the hypothesis. If it is confirmed, promote the concern to a KnownIssue with the observed behavior and evidence. If the hypothesis is unsupported, retain the reasoning behind that disposition.

Resolve the finding

1. Read the concern’s evidence and reproduce or otherwise investigate the suspected behavior. Distinguish absence of a reproduction from evidence that the concern is impossible.
2. Promote a confirmed issue into the KnownIssue workflow, or waive the concern with a substantive reason. The waiver command requires a reason of at least 32 characters.
3. If more investigation is justified, record a realistic review date and the remaining question. Rerun the check and verify that the downstream issue retains the useful evidence.

What a clear result means

A clear result means no inspected open concern is missing its date or due for adjudication. It does not mean the software is free of the suspected failure, that every concern has been discovered or that an investigation established class-wide safety. Waiving an investigation concern is not automatically authorization to ship a known product risk.

Continue with one real issue

Known issues reviewed · Software problem report

Bring one finding, its requirement and the current evidence to scope a walkthrough of what your reviewer can establish and what still needs work.