feat(review): capture fired signals + reversal mapping for the slop and quality gate scores

[RealDiligent]

Summary

• Closes calibration: capture writers + corpus mappings for the slop and quality gate scores #8223
• slopGateMinScore and qualityGateMinScore gate real verdicts today but recorded no signal.rule_fired events, so no corpus could ever form for them and the knob registry could never govern their thresholds.
• Mirrors calibration: wire linked_issue_scope_mismatch into the shared signal-tracking module (fired + reversal) #8101/calibration: wire recordRuleFired/recordHumanOverride for the remaining isConfiguredGateBlocker codes (excludes linked_issue_scope_mismatch, covered by #8101) #8104's capture wiring exactly (the issue's required pattern): recordGateScoreSignals(env, policy, repoFullName, prNumber) fires one rule per knob (slop_gate_score, quality_gate_score) at the env-bearing gate path — wired immediately beside recordConfiguredGateBlockerSignals's existing call — with the same filter each pure evaluation applies: slop evaluates only in block mode with a non-null risk (mirroring buildSlopGateBlocker's own arm, including the default-60 threshold); quality evaluates whenever its mode is not off with both a score and a threshold present (mirroring buildQualityGateWarning's arm), pass AND fail alike — a threshold backtest needs both outcomes. targetKey = repo#pr.
• Metadata carries the score normalized to [0, 1] (both scores are 0-100 integers per normalizeScore, divided by 100 to be confidence-equivalent for buildConfidenceThresholdClassifier replays — the normalization is documented beside the writer, per the issue) plus the detection's own detail string as rawSignal — never diff content, per calibration: capture bounded raw context for the remaining isConfiguredGateBlocker codes, excluding secret_leak #8130's raw-context audit posture for computed-score rules. Best-effort writes (.catch(() => undefined)) that can never affect the verdict.
• Reversal labeling: both ids join CONFIGURED_GATE_BLOCKER_SIGNAL_CODES via a new exported GATE_SCORE_SIGNAL_CODES constant, with the justification the issue requires recorded beside it: slop carries direct gate authority in block mode; quality is advisory-only but registry-governable, and a reversal labels the overall bot outcome its score contributed to — exactly the corpus label the drift/loosening evaluators consume. The fired side is unaffected (that list feeds only the reversal lookups).

Scope

• The PR title follows type(scope): short summary Conventional Commit format, for example fix(api): restore profile access checks.
• This PR is focused and does not mix unrelated backend, UI, MCP, docs, dependency, and deploy changes.
• This follows CONTRIBUTING.md and does not reintroduce GitHub Pages, VitePress, site/, or CNAME.
• I linked a currently open issue this PR resolves (e.g. Closes #123) — a linked open issue is required for every contributor PR.

Validation

• git diff --check
• npm run typecheck
• npm run actionlint
• npm run test:coverage locally; codecov/patch requires ≥99% coverage of the lines AND branches you changed (aim for 100% on your diff so CI variance does not fail near the threshold). Global coverage is a non-blocking trend with a loose 90% backstop, not the gate.
• npm run test:workers
• npm run build:mcp
• npm run test:mcp-pack
• npm run ui:openapi:check
• npm run ui:lint
• npm run ui:typecheck
• npm run ui:build
• npm audit --audit-level=moderate
• New or changed behavior has unit/integration tests for new branches, fallback paths, and sanitizer boundaries

If any required check was skipped, explain why:

• Ran the writer's full suite (vitest run test/unit/configured-gate-blocker-signals.test.ts — 20 tests green, including the 6 new calibration: capture writers + corpus mappings for the slop and quality gate scores #8223 cases: both firing arms per knob with crossing and non-crossing outcomes plus the default slop threshold; every never-evaluated arm (off/advisory-slop modes, null score/threshold); the normalization bounds; the no-diff-content invariant; and the rejecting-SignalStore fail-open path) plus the full root npm run typecheck. actionlint/workers/mcp/ui checks are untouched surfaces; CI runs them all.

Safety

• No secrets, wallet details, hotkeys, coldkeys, user PATs, private keys, raw trust scores, private rankings, or private maintainer evidence are exposed.
• Public GitHub text stays sanitized, low-noise, and does not imply compensation guarantees or optimization tactics.
• Auth, cookie, CORS, GitHub App, Cloudflare, or session changes include negative-path tests.
• API/OpenAPI/MCP behavior is updated and tested where needed.
• UI changes use live API data or real empty/error/loading states, not production mock/demo fallbacks.
• Visible UI changes include a UI Evidence section below with JPG/JPEG or PNG screenshots arranged as organized, captioned, clickable thumbnails. SVG screenshots are not used as review evidence. Review-only screenshots or recordings are not committed to the repository.
• Public docs/changelogs are updated where needed; changelogs are only edited for release-prep PRs.

UI Evidence

Not applicable — backend capture wiring only (no UI, docs, or extension surface touched; no registry entries, no knob movement, per the issue's Boundaries).

Notes

• The corpus becomes visible in the trend automatically once events flow — the rule-row surfaces key off recorded event types, no additional wiring needed (the issue's second deliverable).

[superagent-security]

Superagent didn't find any vulnerabilities or security issues in this PR.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 90.39%. Comparing base (0c8d3d6) to head (3d2ed91).
⚠️ Report is 3 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #8235      +/-   ##
==========================================
- Coverage   92.10%   90.39%   -1.71%     
==========================================
  Files         776       99     -677     
  Lines       78328    26103   -52225     
  Branches    23668     5127   -18541     
==========================================
- Hits        72147    23597   -48550     
+ Misses       5062     2233    -2829     
+ Partials     1119      273     -846     

Flag	Coverage Δ
shard-1	79.08% <66.66%> (+24.16%)	⬆️
shard-2	35.33% <57.14%> (-17.27%)	⬇️
shard-3	40.96% <100.00%> (-13.80%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
src/queue/processors.ts	95.73% <100.00%> (+<0.01%)	⬆️
src/rules/advisory.ts	98.09% <100.00%> (+0.12%)	⬆️

... and 677 files with indirect coverage changes

[loopover-orb]

Caution

🛑 LoopOver review result - reject/close recommended

_{Review updated: 2026-07-23 13:56:04 UTC}

3 files · 1 AI reviewer · 1 blocker · CI green · clean

🛑 Suggested Action - Reject/Close

Review summary
The AI review returned blocking findings for this change but did not include a separate narrative summary. Review the blockers below before deciding this PR.

Blockers

• src/rules/advisory.ts: `recordGateScoreSignals` reads `policy.slopGateMode`/`policy.slopGateMinScore` directly without first calling `applyMergeReadinessGate(policy)` (as `recordConfiguredGateBlockerSignals` does two functions above it), so when a repo sets `mergeReadinessGateMode: block` with `slopGateMode` left unset/advisory, the gate genuinely evaluates and can block on slop via the composite override (since `buildSlopGateBlocker` only ever reads `policy.slopGateMode` directly and has no knowledge of the composite itself), but this function's `slopMode === "block"` check sees the un-overridden mode and skips the write — breaking the PR's own stated invariant that this uses 'the same filter each pure evaluation applies' and silently dropping corpus evidence for exactly the case calibration: capture writers + corpus mappings for the slop and quality gate scores #8223 is meant to capture.

Nits — 5 non-blocking

• The new test file's import reformatting (`import { recordGateScoreSignals,\n RAW_CONTEXT_MAX_DIFF_CHARS, ...`) is stylistically inconsistent with the rest of the file's multi-line import formatting.
• Consider naming `8223` used only in comments with a shared reference (e.g. linking to the issue in one place) rather than repeating the raw number across several doc comments, per the external size/magic-number note.
• No test exercises the `mergeReadinessGateMode: block` + unset `slopGateMode` interaction for `recordGateScoreSignals`, which is exactly the scenario the missing `applyMergeReadinessGate` call would have caught.
• Fix the blocker by computing `const effective = applyMergeReadinessGate(policy);` at the top of `recordGateScoreSignals` and reading `effective.slopGateMode`/`effective.slopGateMinScore` (quality fields can stay off `policy` since readiness is excluded from the composite, per the existing doc comment on `applyMergeReadinessGate`).
• Add a test mirroring `recordConfiguredGateBlockerSignals`'s composite-aware behavior: `mergeReadinessGateMode: "block"` with `slopGateMode` unset should still fire `slop_gate_score`.

Why this is blocked

• src/rules/advisory.ts: `recordGateScoreSignals` reads `policy.slopGateMode`/`policy.slopGateMinScore` directly without first calling `applyMergeReadinessGate(policy)` (as `recordConfiguredGateBlockerSignals` does two functions above it), so when a repo sets `mergeReadinessGateMode: block` with `slopGateMode` left unset/advisory, the gate genuinely evaluates and can block on slop via the composite override (since `buildSlopGateBlocker` only ever reads `policy.slopGateMode` directly and has no knowledge of the composite itself), but this function's `slopMode === "block"` check sees the un-overridden mode and skips the write — breaking the PR's own stated invariant that this uses 'the same filter each pure evaluation applies' and silently dropping corpus evidence for exactly the case calibration: capture writers + corpus mappings for the slop and quality gate scores #8223 is meant to capture.

📋 Copy for AI agents — paste into your coding agent

Fix the following blocker(s) from this PR review:

1. src/rules/advisory.ts: \`recordGateScoreSignals\` reads \`policy.slopGateMode\`/\`policy.slopGateMinScore\` directly without first calling \`applyMergeReadinessGate\(policy\)\` \(as \`recordConfiguredGateBlockerSignals\` does two functions above it\), so when a repo sets \`mergeReadinessGateMode: block\` with \`slopGateMode\` left unset/advisory, the gate genuinely evaluates and can block on slop via the composite override \(since \`buildSlopGateBlocker\` only ever reads \`policy.slopGateMode\` directly and has no knowledge of the composite itself\), but this function's \`slopMode === "block"\` check sees the un-overridden mode and skips the write — breaking the PR's own stated invariant that this uses 'the same filter each pure evaluation applies' and silently dropping corpus evidence for exactly the case \#8223 is meant to capture.

Decision drivers

• ❌ Code review — 1 blocker (1 reviewer)
• ❌ Gate result — Blocking (Repo-configured hard blocker found.)

Context & advisory signals — never blocks the verdict

Signal	Result	Evidence
Linked issue	✅ Linked	#8223
Related work	✅ No active overlap found	No same-issue or scoped active PR overlap found.
Change scope	✅ 20/20	Low review scope from cached public metadata (1 linked issue).
Validation posture	✅ 25/25	PR body includes validation/test evidence.
Contributor workload	✅ 10/10	Author activity: 343 registered-repo PR(s), 136 merged, 36 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	RealDiligent; Gittensor profile; 343 PR(s), 36 issue(s).
Improvement	✅ Minor	risk: clean · value: minor · LLM: moderate

Linked issue satisfaction

Addressed
The PR adds recordGateScoreSignals mirroring the required pattern (createSignalStore, same filters as the pure evaluations, normalized [0,1] confidence documented beside the writer, detail string as rawSignal not diff, best-effort catch), wires it beside recordConfiguredGateBlockerSignals at the env-bearing gate path, and extends the reversal list via a new GATE_SCORE_SIGNAL_CODES constant with ju

Review context

• Author: RealDiligent
• Role context: outside_contributor
• Public audience mode: oss maintainer
• Lane context: Repository is configured for direct PR review.
• Public profile languages: Python, Ruby, JavaScript, Svelte, TypeScript, Markdown, MDX, Rust
• Official Gittensor activity: 343 PR(s), 36 issue(s).
• PR-specific overlap: none found.

Contributor next steps

• Keep the PR focused and include validation evidence before maintainer review.

Signal definitions

• Related work = same linked issue, overlapping active PRs, or title/path similarity.
• Change scope = cached public metadata such as size labels, draft state, and review-burden hints.
• Validation posture = whether the PR provides enough public validation/test evidence for maintainer review.
• Contributor workload = public contributor activity and cleanup pressure, not a repo-wide quality failure.
• Contributor context = public GitHub/Gittensor identity context; non-Gittensor status is not a blocker.

🧪 Chat with LoopOver

Ask LoopOver a question about this PR directly in a comment — grounded only in the same cached, public-safe facts shown above, never a new claim.

• @loopover ask <question> answers contribution-quality Q&A with source citations and freshness.
• @loopover chat <question> answers in natural prose from cached decision-pack facts via local inference (maintainer/collaborator; read-only).
• A plain-language @loopover mention with a real question is routed to the closest matching read-only command automatically — no exact syntax required.

Full command reference: https://loopover.ai/docs/loopover-commands

🧪 Experimental — new and may change.

_{🟩 Safe / merged · 🟦 Advisory · 🟨 Held for review · 🟥 Blocked / closed}

---

💰 Earn for open-source contributions like this. Gittensor lets GitHub contributors earn for the work they already do — register to start earning →.

Checked by LoopOver, a quiet PR intelligence layer for OSS maintainers.

• Re-run LoopOver review

[loopover-orb]

LoopOver is closing this pull request on the maintainer's behalf (AI reviewers agree on a likely critical defect: src/rules/advisory.ts: `recordGateScoreSignals` reads `policy.slopGateMode`/`policy.slopGateMinScore` directly without first calling `applyMergeReadinessGate(policy)` (as `recordConfiguredGateBlockerSignals` does two functions above it), so when a repo sets `mergeReadinessGateMode: block` with `slopGateMode` left unset/advisory, the gate genuinely evaluates and can block on slop via the composite override (since `buildSlopGateBlocker` only ever reads `policy.slopGateMode` directly and has no knowledge of the composite itself), but this function's `slopMode === "block"` check sees the un-overridden mode and skips the write — breaking the PR's own stated invariant that this uses 'the same filter each pure evaluation applies' and silently dropping corpus evidence for exactly the case #8223 is meant to capture.). This is an automated maintenance action — to pursue this change, please open a new pull request with the issues resolved. Closed PRs may be analyzed later to improve review accuracy, but they are not automatically reopened or re-reviewed.