fix(chat): redact private lane signals from Q&A

[JSONbored]

Motivation

• Prevent private reward/lane economics (e.g. Maintainer cut, direct PR lane share, split lane, issue-discovery) from being included in the LLM grounding or published in public PR comments via the @gittensory chat Q&A path.

Description

• Add a PRIVATE_LANE_SIGNAL_PATTERN to detect lane/reward phrasing and redact those matches from grounding text before building the LLM prompt in src/services/ai-chat-qa.ts.
• Extend the model-output forbidden-text guard (PUBLIC_FORBIDDEN_TEXT_PATTERN) to treat these lane-signal phrases as forbidden so generated answers that repeat them are withheld.
• Wire the new pattern into redactGroundingText so action.why lines containing lane signals are replaced with private readiness context before prompting the provider.
• Add regression tests in test/unit/ai-chat-qa.test.ts to assert prompt redaction, output rejection when the provider repeats lane signals, and helper-level sanitization behavior.

Testing

• Ran npx vitest run test/unit/ai-chat-qa.test.ts and the suite passed (29 tests) against the modified code.
• Ran npm run typecheck and tsc --noEmit which completed successfully on the changes.
• Ran scoped coverage for the file (npm run test:coverage targeted at the unit file) which passed for the modified tests, but a full repository coverage run still fails the global thresholds (the repo-wide coverage requirement is unchanged and requires broader test coverage than this focused fix).
• Attempting the full gate npm run test:ci failed at cf-typegen:check due to stale generated worker types (the repo reports worker-configuration.d.ts needs regeneration) and npm audit --audit-level=moderate could not complete due to an audit endpoint 403 in this environment.

---

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 94.14%. Comparing base (cd1a905) to head (b89c02b).
⚠️ Report is 56 commits behind head on main.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #5149   +/-   ##
=======================================
  Coverage   94.13%   94.14%           
=======================================
  Files         466      466           
  Lines       39573    39574    +1     
  Branches    14436    14436           
=======================================
+ Hits        37254    37255    +1     
  Misses       1664     1664           
  Partials      655      655           

Flag	Coverage Δ
shard-1	46.50% <100.00%> (-0.02%)	⬇️
shard-2	33.44% <100.00%> (-0.12%)	⬇️
shard-3	31.60% <100.00%> (+0.25%)	⬆️
shard-4	32.70% <100.00%> (+0.01%)	⬆️
shard-5	33.19% <100.00%> (-0.24%)	⬇️
shard-6	45.14% <100.00%> (+0.20%)	⬆️

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

Files with missing lines	Coverage Δ
src/services/ai-chat-qa.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: 2026-07-12 09:01:29 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 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/services/ai-chat-qa.ts:68 leaves bare `direct PR` out of `PUBLIC_FORBIDDEN_TEXT_PATTERN` even though src/services/ai-chat-qa.ts:63 redacts it as private grounding, so a provider response like `direct PR 1 is the useful lane` returns `ok` instead of `unsafe`; add `direct PR` to the output forbidden pattern or remove it from the private lane signal list with a reason.

Nits — 5 non-blocking

• The bare `direct PR` alternative in PRIVATE_LANE_SIGNAL_PATTERN (src/services/ai-chat-qa.ts:60-61) will redact any grounding text mentioning a plain direct PR (unrelated to lane economics), which is broader than the stated intent — consider requiring `lane`/`share` context to avoid over-redacting benign phrasing.
• PUBLIC_FORBIDDEN_TEXT_PATTERN (src/services/ai-chat-qa.ts:66) omits the bare `direct PR` term that PRIVATE_LANE_SIGNAL_PATTERN redacts on input, so a model that independently generates the bare phrase `direct PR` (not from grounding) would not be caught by containsPublicForbiddenText — worth confirming this asymmetry is intentional.
• The PR description doesn't reference an open issue number (e.g. `Fixes #NNNN`) that this change closes, which this repo's contribution policy requires for external PRs.
• Consider tightening the `direct PR` alternative to require accompanying lane/share/percentage context so plain mentions of a direct PR aren't unnecessarily rewritten to "private readiness context".
• Add a code comment cross-referencing PRIVATE_LANE_SIGNAL_PATTERN and the lane-related terms folded into PUBLIC_FORBIDDEN_TEXT_PATTERN so future edits keep the two patterns in sync (mirroring the existing comment style for PRIVATE_DECISION_BLOCKER_PATTERN).

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: 45 registered-repo PR(s), 37 merged, 409 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	JSONbored; Gittensor profile; 45 PR(s), 409 issue(s).
Gate result	❌ Blocking	Repo-configured hard blocker found.
Improvement	✅ Minor	risk: clean · value: minor — Code changes are accompanied by test evidence.

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: 45 PR(s), 409 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.

[BETA] Chat with Gittensory

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

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

Full command reference: https://gittensory.aethereal.dev/docs/gittensory-commands

_{🟩 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