feat(github-app): add reviewer routing auto-request mode

[statxc]

Summary

• Implements feat(github-app): reviewer-routing auto_request action (#540 follow-up) #830 by adding CODEOWNERS-based reviewer routing with reviewerRoutingMode: off | advisory | auto_request
• Surfaces reviewer suggestions in the PR review output
• Auto-requests top-ranked individual reviewers only when reviewerRoutingMode is auto_request, while skipping teams, already-requested reviewers, and first-time external contributors
• Adds the required settings, API/OpenAPI plumbing, migration, and focused tests for routing, idempotency, newcomer guards, and gate-override safety
• Includes the follow-up CODEOWNERS matcher fix so directory and shallow-wildcard rules align with GitHub's documented semantics

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 an issue, or this is small enough that the summary explains why an issue is not needed.

Validation

• git diff --check
• npm run actionlint
• npm run typecheck
• npm run test:coverage
• npm run test:workers
• npm run build:mcp
• npm run test:mcp-pack
• npm run ui:openapi:check
• npm run ui:version-audit
• npm run ui:lint
• npm run ui:typecheck
• npm run ui:test
• 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 or failed, explain why:

• I ran npm run test:ci locally against the final PR tree.
• ui:lint passes, but it emits existing react-refresh/only-export-components warnings in route files outside this PR's scope.

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. This PR does not introduce visible UI changes. The frontend-facing change is the generated OpenAPI artifact only.

State / title	JPG/PNG evidence
N/A	No visible UI change

Notes

• Closes feat(github-app): reviewer-routing auto_request action (#540 follow-up) #830
• This PR includes the CODEOWNERS routing foundation required to implement the auto_request reviewer-routing mode correctly on current main.
• auto_request is intentionally conservative: it only requests individual reviewers, never teams, and it does not auto-request for first-time external contributors.
• The branch is intentionally squashed to a single commit for merge review.

[codecov]

Codecov Report

❌ Patch coverage is 98.23789% with 4 lines in your changes missing coverage. Please review.
✅ Project coverage is 95.07%. Comparing base (a6e32e7) to head (b3e8364).
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
src/github/codeowners.ts	97.72%	0 Missing and 2 partials ⚠️
src/queue/processors.ts	95.00%	0 Missing and 1 partial ⚠️
src/signals/reviewer-routing.ts	98.46%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1108      +/-   ##
==========================================
+ Coverage   95.03%   95.07%   +0.04%     
==========================================
  Files         177      179       +2     
  Lines       19886    20101     +215     
  Branches     7166     7244      +78     
==========================================
+ Hits        18898    19112     +214     
+ Misses        395      393       -2     
- Partials      593      596       +3     

Files with missing lines	Coverage Δ
src/api/routes.ts	94.40% <ø> (ø)
src/db/repositories.ts	96.06% <100.00%> (+<0.01%)	⬆️
src/db/schema.ts	68.32% <ø> (ø)
src/github/app.ts	97.52% <100.00%> (+0.58%)	⬆️
src/github/backfill.ts	92.86% <100.00%> (+0.03%)	⬆️
src/openapi/schemas.ts	100.00% <ø> (ø)
src/signals/engine.ts	97.07% <100.00%> (+0.01%)	⬆️
src/signals/focus-manifest-loader.ts	98.87% <ø> (ø)
src/signals/focus-manifest.ts	96.57% <100.00%> (+0.01%)	⬆️
src/signals/settings-preview.ts	99.24% <100.00%> (+0.01%)	⬆️
... and 3 more

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[ghost]

Tip

🟩🟩🟩🟩🟩🟩🟩🟩🟩🟩🟩🟩

✅ Gittensory review — safe to merge

37 files · no blockers · readiness 100/100 · CI green

✅ Approved — safe to merge

Signal	Result	Evidence
Code review	✅ No blockers	synthesized
Linked issue	✅ Linked	#830
Related work	✅ No active overlap found	No same-issue or scoped active PR overlap found.
Review load	✅ 20/20	Readiness component derived from cached public PR metadata and labels; size label size:XL.
Validation evidence	✅ 25/25	PR body includes validation/test evidence.
Open PR queue	✅ 10/10	0 open PR(s), 0 likely reviewable.
Contributor context	✅ Confirmed Gittensor contributor	statxc; Gittensor profile; 46 PR(s), 4 issue(s).
Gate result	✅ Passing	No configured blocker found.

Review context

• Author: statxc
• Role context: outside_contributor
• Public audience mode: oss maintainer
• Lane context: Repository is configured for direct PR review.
• Public profile languages: TypeScript, Python, JavaScript, Rust, C++, Go, Java, Clojure
• Official Gittensor activity: 46 PR(s), 4 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.
• Review load = cached public PR metadata such as size labels, changed paths, and preflight status.
• Open PR queue = repo-wide review pressure; it is not a PR quality failure.
• Contributor context = public GitHub/Gittensor identity context; non-Gittensor status is not a blocker.

Visual preview

Route	Viewport	Before (production)	After (this PR's preview)
/	desktop
/	mobile

Click any thumbnail to open the full-size screenshot. Before = production · After = this PR's preview deploy.

_{🟩 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 Gittensory, a quiet PR intelligence layer for OSS maintainers.

• Re-run Gittensory review

[superagent-security]

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

[statxc]

@JSONbored could you please review this PR?

[JSONbored]

Nice work here — the CODEOWNERS parser, the load-de-weighted ranking, and the test coverage are
all solid. I'm closing rather than merging for two reasons: this is two features in one PR, and
the auto_request write has a real safety hole. Resubmit as the stack at the bottom and it's in
good shape.

1. Scope — two features in one PR. The advisory routing core (src/github/codeowners.ts +
   src/signals/reviewer-routing.ts, both new) stands on its own and doesn't need the write path
   at all. #830 itself says the auto_request action "warrants its own focused review" because it
   acts on live PRs. Land them separately (stack below).

2. auto_request fires under dry-run — this is the blocker. Trace:

   • maybeAutoRequestReviewers (src/queue/processors.ts:1879) has no action-mode check — only
     the reviewerRoutingMode === "auto_request" opt-in.
   • it's called unconditionally at src/queue/processors.ts:2545, inside
     maybePublishPrPublicSurface, which still RUNS when a repo is in dry-run (it computes the
     gate verdict; only its Octokit writes get suppressed).
   • requestPullRequestReviewers (src/github/app.ts) writes with a raw fetch, so it never
     goes through the dry-run chokepoint makeInstallationOctokit (src/github/client.ts).
   • Net: a repo set to dry-run still requests real reviewers on real PRs.

   We routed every installation write through the dry-run-aware client in #1258 specifically to
   stop exactly this, and retrofitted the remaining raw-fetch writers in #1263 to honor the brake.
   Follow that pattern — preferably call through makeInstallationOctokit so dry-run/pause suppress
   it for free; or, if it stays a raw-fetch helper, early-return like fileUpstreamDriftIssues:

   // a paused / frozen / dry-run agent must not write (mirror #1263)
   if (isGlobalAgentPause(env) || (await isGlobalAgentFrozen(env))) return;

   Then confirm there's no other path that reaches the write under a non-live mode. Keep the
   newcomer opt-in and the already-requested idempotency you have — those are good.

Suggested stack (smallest first; split further if any exceeds ~10 files):
PR 1 — src/github/codeowners.ts (parser, pure) + test
PR 2 — src/signals/reviewer-routing.ts (ranking) + test
PR 3 — reviewerRoutingMode off|advisory: types/db/schema + migration + regenerated
openapi.json + settings-preview + focus-manifest + the PR-panel "Suggested reviewers"
render + the one-line fixture updates
PR 4 — auto_request only, with the dry-run fix above + tests

PRs 1–3 are low-risk and move fast; PR 4 then gets the focused review the live write needs.

Per PR: rebase onto latest main, 100% line + branch coverage with a regression test, no
AI-generated attribution in the PR body.