fix(selfhost): canonicalize repo-doc refresh repos

[JSONbored]

Motivation

• Prevent mixed-case owner/repo names from bypassing per-repository pause/dry-run settings when performing on-demand repo-doc refreshes by ensuring the stored repository canonical name is used for settings and write operations.

Description

• Resolve the stored repository record with getRepository and derive a canonicalRepoFullName before calling getRepositorySettings, openRepoDocPullRequest, or recording attempt snapshots in performRepoDocRefresh (src/github/repo-doc-refresh-runner.ts).
• Import getRepository alongside existing DB helpers and use the canonical name for settings lookup, PR-opening, and attempt recording to avoid casing mismatches.
• Add regression tests in test/unit/repo-doc-refresh-runner.test.ts that cover the uninstalled-repo fallback and a mixed-case caller path that must respect the stored agentPaused setting.

Testing

• Ran npm run typecheck which succeeded with no TypeScript errors.
• Ran the targeted unit tests with npx vitest run test/unit/repo-doc-refresh-runner.test.ts and all tests in that file passed.
• Attempted the full local gate with npm run test:ci, but the run was interrupted/blocked by unrelated long-running failures/timeouts in existing test suites (test/unit/queue.test.ts and test/unit/backfill.test.ts) in the local environment, so the full gate was not completed here.

---

Codex Task

[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 93.11%. Comparing base (15ace01) to head (a69f0c9).
⚠️ Report is 24 commits behind head on main.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #3605   +/-   ##
=======================================
  Coverage   93.11%   93.11%           
=======================================
  Files         305      305           
  Lines       31532    31534    +2     
  Branches    11509    11510    +1     
=======================================
+ Hits        29361    29363    +2     
  Misses       1517     1517           
  Partials      654      654           

Files with missing lines	Coverage Δ
src/github/repo-doc-refresh-runner.ts	100.00% <100.00%> (ø)

🚀 New features to boost your workflow:

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

[loopover-orb]

Warning

🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨

⏸️ Gittensory review result - manual review recommended

_{Review updated: 2026-07-05 19:39:39 UTC}

2 files · 1 AI reviewer · 2 blockers · readiness 93/100 · CI green · clean

⏸️ Suggested Action - Manual Review

• No linked issue detected — If this PR is intended to solve an issue, link it explicitly in the PR body.
• Maintainer requires a linked issue — Link the relevant issue (for example Closes #123) before opening the PR.

Review summary
This fix resolves getRepository(env, repoFullName) once in performRepoDocRefresh and threads the resulting canonical fullName through getRepositorySettings, openRepoDocPullRequest, and recordRepoDocRefreshAttempt, closing a real bypass where a mixed-case caller (e.g. the MCP on-demand trigger) could miss a stored agentPaused/settings row keyed by the canonical casing. The new mixed-case test (test/unit/repo-doc-refresh-runner.test.ts) actually drives the real path — it seeds settings under lowercase 'owner/widgets', calls performRepoDocRefresh with 'Owner/Widgets', and asserts the paused mode is honored and the attempt is recorded only under the canonical name — so this isn't a fabricated-state test, and CI is green. The uninstalled-repo fallback (repository is null) correctly falls back to the raw repoFullName for both settings lookup and attempt recording, which is the same pre-fix behavior for that case.

Nits — 5 non-blocking

• src/github/repo-doc-refresh-runner.ts:59-62 adds one extra getRepository DB round-trip per refresh call — negligible for this single-repo, on-demand path but worth noting if this function is ever called in a hot loop.
• The PR description doesn't cite a specific tracked issue number this closes, only referencing the pre-existing feat(selfhost): scheduled + on-demand refresh for repo-doc generation #3003 code comment — worth confirming there's an eligible open issue this bypass fix is linked to per repo contribution policy.
• test/unit/repo-doc-refresh-runner.test.ts's new mixed-case test asserts getLastRepoDocRefreshAttemptedAt(env, 'Owner/Widgets') is null but doesn't also assert the openRepoDocPullRequest call itself received the canonical name (it infers this indirectly via the paused-mode outcome) — an explicit spy/assert on the call argument would make the canonicalization guarantee more directly verifiable.
• Consider adding a short code comment at src/github/repo-doc-refresh-runner.ts:59 noting why canonicalRepoFullName falls back to the raw repoFullName (uninstalled repo) so future readers don't assume getRepository always succeeds.
• If getRepository's name lookup isn't already case-insensitive at the DB layer, confirm that here (not shown in this diff) since the whole fix depends on it resolving 'Owner/Widgets' to the stored 'owner/widgets' row.

Concerns raised — review before merging

• No linked issue detected — If this PR is intended to solve an issue, link it explicitly in the PR body.
• Maintainer requires a linked issue — Link the relevant issue (for example Closes #123) before opening the PR.

Signal	Result	Evidence
Code review	❌ 2 blockers	1 reviewer
Linked issue	⚠️ Missing	No linked issue or no-issue rationale found.
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 (no linked issue context).
Validation posture	✅ 25/25	PR body includes validation/test evidence.
Contributor workload	✅ 10/10	Author activity: 56 registered-repo PR(s), 46 merged, 493 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	JSONbored; Gittensor profile; 56 PR(s), 493 issue(s).
Gate result	❌ Blocking	Repo-configured hard blocker found.

Review context

• Author: JSONbored
• Role context: owner (maintainer lane)
• Public audience mode: oss maintainer
• Lane context: Repository registration is not available in the local Gittensory cache.
• Public profile languages: Python, TypeScript, JavaScript, Ruby, Go, Kotlin, MDX, Shell
• Official Gittensor activity: 56 PR(s), 493 issue(s).
• PR-specific overlap: none found.

Contributor next steps

• Treat this as maintainer-lane context rather than normal contributor-lane activity.
• Explain no-issue PR.
• No action.
• Link the issue being solved, or explicitly explain why this is a no-issue PR.

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.

_{🟩 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