fix(review): make review.selftune: false opt-out absolute for the breaker too

[JSONbored]

Summary

• The accuracy circuit-breaker (runSelfTuneBreaker) had no per-repo opt-out at all, unlike its sibling selfTuneRepos() (the routine tuning pass), which already correctly excludes a repo whose .loopover.yml sets review.selftune: false. A repo could opt out of routine tuning yet still have its gate mode forced into holdonly/closehold by the breaker.
• Per that flag's own documented intent ("excludes this repo from the tuning pass"), the opt-out is now absolute: an opted-out repo is excluded from every part of self-tune, both the merge and close breakers, in both the plain and miner-scoped (maintainer: wire calibration accuracy into the live auto-tune circuit-breaker for miner-originated PRs #2352) passes, and in both directions — the breaker can neither newly engage a hold for an opted-out repo nor auto-clear an already-engaged one. A manifest-load error fails open (repo stays included), matching selfTuneRepos()'s own fail-safe precedent.

Closes #6803

Test plan

• npm run test:ci — green for every check touching this diff (one unrelated, pre-existing failure on main — a stale hardcoded gittensory repo name in an unrelated test — flagged separately)
• 4 new regression tests: engage-blocked, auto-clear-blocked (absolute in both directions), miner-scoped pass, and a control case confirming the exclusion is per-repo, not global

[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 93.72%. Comparing base (018e261) to head (878e8a0).
⚠️ Report is 2 commits behind head on main.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6995      +/-   ##
==========================================
+ Coverage   89.25%   93.72%   +4.46%     
==========================================
  Files         688      688              
  Lines       68491    68506      +15     
  Branches    18737    18740       +3     
==========================================
+ Hits        61135    64207    +3072     
+ Misses       5647     3302    -2345     
+ Partials     1709      997     -712     

Flag	Coverage Δ
shard-1	43.25% <0.00%> (-0.42%)	⬇️
shard-2	36.87% <0.00%> (+<0.01%)	⬆️
shard-3	32.89% <41.17%> (-0.13%)	⬇️
shard-4	34.41% <100.00%> (?)
shard-5	31.72% <0.00%> (-0.40%)	⬇️
shard-6	45.95% <0.00%> (+0.09%)	⬆️

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

Files with missing lines	Coverage Δ
src/review/outcomes-wire.ts	90.96% <100.00%> (+43.28%)	⬆️

... and 99 files with indirect coverage changes

[loopover-orb]

Tip

✅ LoopOver review result - approve/merge recommended

_{Review updated: 2026-07-17 17:48:00 UTC}

2 files · 1 AI reviewer · no blockers · readiness 100/100 · CI green · clean

✅ Suggested Action - Approve/Merge

• safe to merge

Review summary
This PR closes a real gap (#6803): the accuracy circuit-breaker previously ignored a repo's `review.selftune: false` opt-out that `selfTuneRepos()` already respected, letting an opted-out repo still get forced into holdonly/closehold. The fix computes exclusions before engage/clear candidates are derived (via `excludeSelfTuneOptedOut`), applies to both plain and miner-scoped passes, works bidirectionally (blocks engage and auto-clear), and fails open on manifest-load error consistent with the sibling routine-tuning path. The change is well-scoped to the stated issue, includes 4 targeted regression tests covering engage-block, clear-block, miner-scoped exclusion, and per-repo isolation, and CI is green across all checks.

Nits — 6 non-blocking

• src/review/outcomes-wire.ts: `isSelfTuneOptedOut` calls `loadRepoFocusManifest` once per candidate project sequentially in a for-loop inside `excludeSelfTuneOptedOut`; for a large candidate set this serializes manifest loads rather than using `Promise.all`, though the set size is bounded by engaged/reported projects per breaker run so it's unlikely to matter in practice.
• test/unit/outcomes-wire.test.ts: the new describe block's tests rely on exact seed counts (e.g. 4/8/4/8 splits) mirroring earlier tests by comment reference only ('Same shape as the plain ENGAGES tests above') rather than a shared helper, which is a bit fragile if those upstream tests' thresholds change.
• src/review/outcomes-wire.ts: doc comments on `baseProjectName`, `isSelfTuneOptedOut`, and `excludeSelfTuneOptedOut` are quite long relative to the code they document; the file was already flagged as long (707 lines) and these verbose comments add to that without much marginal clarity beyond the PR description.
• Consider `Promise.all(Array.from(candidateProjects, checkOptOut))` in `excludeSelfTuneOptedOut` to parallelize manifest loads across projects.
• A brief top-of-block comment referencing the shared seeding pattern (or extracting a small helper) would reduce duplication risk between the new opt-out tests and their earlier ENGAGES/AUTO-CLEARS counterparts.
• PR author also opened the linked issue — Link an issue that was opened by a different contributor, or provide a rationale for why this self-authored issue represents genuine discovery work.

Decision drivers

• ✅ Code review — No blockers (1 reviewer)
• ✅ Gate result — Passing (No configured blocker found.)

Context & advisory signals — never blocks the verdict

Signal	Result	Evidence
Linked issue	✅ Linked	#6803
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: 33 registered-repo PR(s), 26 merged, 341 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	JSONbored; Gittensor profile; 33 PR(s), 341 issue(s).
Improvement	✅ Minor	risk: clean · value: minor · LLM: moderate

Linked issue satisfaction

Addressed
The PR implements the 'yes' deliverable by threading the review.selftune:false opt-out into runSelfTuneBreaker via excludeSelfTuneOptedOut, filtering both the plain and miner-scoped passes in both engage and clear directions, with a fail-open manifest-load fallback matching selfTuneRepos()'s precedent.

Review context

• Author: JSONbored
• Role context: owner (maintainer lane)
• Public audience mode: oss maintainer
• Lane context: Repository is configured for direct PR review.
• Public profile languages: not available
• Official Gittensor activity: 33 PR(s), 341 issue(s).
• PR-specific overlap: none found.

Contributor next steps

• Start here: Treat this as maintainer-lane context rather than normal contributor-lane activity.

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 approves — the gate is satisfied and CI is green.