fix(rag): scope chunk reads by repository

[JSONbored]

Motivation

• Prevent a cross-repository RAG metadata leak where vector matches from a truncated shared namespace could expose another repo's retrieved paths and text when only the requested repo was authorized.

Description

• Scope chunk-text lookups by project and repo so vector matches can only hydrate text when the stored repo_chunks row belongs to the authorized repository by changing retrieveContextWithMetrics to call readChunkTexts(storage, project, repo, ids).
• Update readChunkTexts to accept project + repo and include WHERE project = ? AND repo = ? AND id IN (...) so D1 reads cannot return victim repo rows for a different requested repo.
• Add a regression unit test that constructs two repo names with the same truncated RAG namespace and asserts that a vector match whose chunk row belongs to the colliding (victim) repo is dropped (no context, kept chunks, or retrieved paths).
• Regenerate Cloudflare worker runtime types after the local cf-typegen check reported drift.

Testing

• Ran npx vitest run test/unit/rag.test.ts, which passed the modified and new tests for retrieveContextWithMetrics and readChunkTexts.
• Ran git diff --check and npm run cf-typegen, both of which succeeded and produced the updated worker-configuration.d.ts artifact.
• Ran npm run test:ci, which failed to complete due to existing timeout failures in unrelated test/unit/queue.test.ts sweep tests during the coverage run; the local coverage run was stopped after those timeouts and is not addressed by this change.
• Ran npm audit --audit-level=moderate, which failed due to the npm registry audit endpoint returning 403 Forbidden (environmental/registry issue, not caused by this change).

---

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 94.22%. Comparing base (3ae7ece) to head (ae4d002).
⚠️ Report is 7 commits behind head on main.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #4651   +/-   ##
=======================================
  Coverage   94.22%   94.22%           
=======================================
  Files         441      441           
  Lines       38741    38741           
  Branches    14117    14117           
=======================================
  Hits        36504    36504           
  Misses       1577     1577           
  Partials      660      660           

Files with missing lines	Coverage Δ
src/review/rag.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: 2026-07-11 00:22:42 UTC}

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

⏸️ 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
The diff correctly scopes `readChunkTexts` by `project` and `repo` via a `WHERE project = ? AND repo = ? AND id IN (...)` clause and updates the sole call site in `retrieveContextWithMetrics` to pass `opts.project`/`opts.repo`, directly closing the cross-repo vector-namespace collision described in the motivation. The new test builds two repo names that collide under `ragNamespace`'s truncation and asserts the victim's chunk row is dropped when the requester's project/repo don't match the bound query params — this exercises the real D1 filtering behavior rather than a fabricated scenario, since the mock only returns rows when the bound values equal the victim's stored key. The bundled `worker-configuration.d.ts` regen is unrelated generated churn but is disclosed as a required `cf-typegen` drift fix.

Nits — 5 non-blocking

• src/review/rag.ts: the PR bundles a large, unrelated `worker-configuration.d.ts` regen with a security fix — consider splitting generated-artifact updates into their own PR going forward for cleaner history.
• The PR description doesn't reference a tracked issue for this fix; even for a security-motivated change, linking the originating report/issue would help future readers understand provenance.
• test/unit/rag.test.ts: consider adding a case where the vector match's chunk row belongs to the *same* repo/project to explicitly assert the positive (non-dropped) path stays intact alongside the new negative-path test.
• Confirm `readChunkTexts`'s catch-path error logging includes enough context (project/repo) to debug scoped-query failures in production, since the WHERE clause is now more specific.
• If other call sites of `readChunkTexts` exist beyond `retrieveContextWithMetrics`, verify they were all updated to the new 4-arg signature (not shown in this diff).

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: 48 registered-repo PR(s), 40 merged, 275 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	JSONbored; Gittensor profile; 48 PR(s), 275 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 is configured for direct PR review.
• Public profile languages: not available
• Official Gittensor activity: 48 PR(s), 275 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.
• 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