Evaluate Findings

Assess external feedback (code reviews, AI suggestions, PR comments) with adversarial verification. Triage findings into actionable verdicts. Do not apply fixes.

Step 1: Assess Each Finding

If you already assessed a finding earlier in this session and recorded a verdict of Skip or Escalate — for example when an iterating loop re-runs review and the same finding resurfaces — do not re-adjudicate it from scratch. When the re-reported finding matches one you already judged (same location and substance) and presents no new evidence beyond what your recorded reason already accounts for, keep that verdict and reason without re-reading the code, re-verifying, or routing it to the Devil's Advocate in Step 2. Assess fresh only when the finding raises materially new evidence, or when you have not judged it before in this session.

When several findings rest on a shared premise — for example a source-of-truth choice — verify that premise once before adjudicating them individually. Findings whose premise holds proceed through normal per-finding verification; when it fails, they are all Skip, citing the refuted premise.

When a plan, spec, or shell file governs the work, re-read the decisions it records before adjudicating. Having read it earlier in the session does not count: once it falls out of context, a recorded decision is indistinguishable from no decision at all.

For each finding: