feat(zero-noise): signal verdicts + published precision on /status #5524
ccantynzAI Reviewcommented 8d ago
AI Triage
(no summary)
Priority: medium Risk area: mixed
Suggested labels: (no label suggestions) Suggested reviewers: (no reviewer suggestions)
Suggestions only — nothing has been applied. The PR author stays in control.
gluecron[bot]🤖 botAI Reviewcommented 8d ago
AI review unavailable
The platform's AI balance is exhausted, so AI generation is temporarily unavailable. Nothing was lost. You can queue this as a repair for the internal agent from the repository's Health page, or try again once the balance is restored. The PR is otherwise unchanged.
Cross-repo impact
See what breaks downstream if this PR merges.
⮌ Merged
This pull request was merged into main.
c comment · e edit title · m merge · a approve · r request changes · ? shortcuts
Zero-Noise build-map item 1 (agent-built, coordinator-rebased)
The number nobody in developer tooling publishes, now computed and public:
pr_comments.is_ai_review— the column every existing surface already means by "AI review comment"; the summary-marker approach was rejected as it misses inline findings). PR author or write access may judge — deliberately narrower than thread-resolve, since verdicts feed a published figure. Migration 0127, unique standing verdict per signal with upsert./status"Signal precision": precision = acted/(acted+dismissed) over rolling 30 days, always rendered beside its sample size, publishing only at ≥10 judgments — below that, the honest empty state with real counts. Every state carries the provenance line: "Computed from developer verdicts on AI review findings, not claimed." Three states including store-unreadable; no path fabricates a number.Zero new env vars. Typecheck clean; 43 tests green across precision/copy-honesty/exclusion-rot (7 DB-gated skips run in CI).
Next per the doctrine's build map: verified-before-shown (the skeptic pass), then Actions compatibility.
🤖 Generated with Claude Code