fix(review): keep recap cohorts private

[JSONbored]

Motivation

• The scheduled maintainer recap was opting loadGatePrecisionReport(..., { includeCohorts: true }) and rendering miner-vs-human cohort diagnostics into external Discord/Slack digests, which exposes sensitive maintainer-review/identity signals on public surfaces.
• The change restores the public/private boundary by keeping cohort diagnostics as a maintainer-only diagnostic and preventing cohort labels/counts from appearing in externally-delivered recap bodies.
• Add defense-in-depth by extending the shared public redaction vocabulary to catch cohort-related labels if they ever surface.

Description

• Stop requesting cohort data for scheduled digests by removing includeCohorts: true at the scheduled loader call in src/review/maintainer-recap-wire.ts so periodic recap deliveries remain aggregate-only.
• Keep the internal totals.cohorts shape available to callers that inject cohort-bearing reports, but remove cohort-derived lines from the public summary and omit the ## Cohorts section from the formatted public body in src/services/maintainer-recap.ts and formatMaintainerRecap.
• Extend PUBLIC_UNSAFE_TERMS in src/signals/redaction.ts to include cohort diagnostic vocabulary (cohort, miner-originated, human-originated) as defense-in-depth against accidental public leakage.
• Update unit tests to assert that cohort diagnostics are not exposed on the public recap path (test/unit/maintainer-recap.test.ts, test/unit/maintainer-recap-format.test.ts, test/unit/maintainer-recap-wire.test.ts).
• Regenerate Cloudflare type output via the local typegen run resulting in an updated worker-configuration.d.ts to keep the type-check/generation gate current.

Testing

• Ran targeted unit tests: npx vitest run test/unit/maintainer-recap.test.ts test/unit/maintainer-recap-format.test.ts test/unit/maintainer-recap-wire.test.ts --reporter=dot and all targeted tests passed (44 tests across 3 files).
• Typecheck: npm run typecheck succeeded with no reported errors.
• Cloudflare types: npm run cf-typegen executed and updated worker-configuration.d.ts to match the environment.
• Full CI: npm run test:ci was attempted but failed in unrelated, existing queue/sweep integration tests (timing/remote setup) in this environment; the maintainer-recap-specific tests used for this fix passed.
• Dependency audit: npm audit --audit-level=moderate could not complete in this environment (registry returned 403), so audit was not verified 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 94.34%. Comparing base (5f3ebf7) to head (7d14161).
⚠️ Report is 2 commits behind head on main.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #5204      +/-   ##
==========================================
- Coverage   94.35%   94.34%   -0.01%     
==========================================
  Files         473      473              
  Lines       39977    39968       -9     
  Branches    14574    14570       -4     
==========================================
- Hits        37719    37708      -11     
  Misses       1585     1585              
- Partials      673      675       +2     

Flag	Coverage Δ
shard-1	46.42% <100.00%> (-0.04%)	⬇️
shard-2	34.58% <100.00%> (+0.02%)	⬆️
shard-3	31.97% <100.00%> (+1.00%)	⬆️
shard-4	32.20% <100.00%> (-1.11%)	⬇️
shard-5	33.69% <100.00%> (+<0.01%)	⬆️
shard-6	44.88% <100.00%> (+0.01%)	⬆️

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

Files with missing lines	Coverage Δ
packages/gittensory-engine/src/prompt-packet.ts	100.00% <100.00%> (ø)
src/review/maintainer-recap-wire.ts	98.07% <ø> (ø)
src/services/maintainer-recap.ts	96.77% <ø> (-3.23%)	⬇️
src/signals/redaction.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]

Caution

🟥🟥🟥🟥🟥🟥🟥🟥🟥🟥🟥🟥

🛑 Gittensory review result - fixes required

_{Review updated: 2026-07-12 09:54:57 UTC}

6 files · 1 AI reviewer · 2 blockers · readiness 93/100 · CI failing · blocked

🛑 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 PR removes the `includeCohorts: true` opt-in at the scheduled maintainer-recap call site (src/review/maintainer-recap-wire.ts) so periodic Discord/Slack digests no longer request miner-vs-human cohort data, strips the cohort summary line and `## Cohorts` section from the public formatter (src/services/maintainer-recap.ts), and adds `cohort`/`miner-originated`/`human-originated` to the shared `PUBLIC_UNSAFE_TERMS` redaction vocabulary as defense-in-depth (src/signals/redaction.ts). The fix is applied at the correct source (the request-time opt-in) rather than only patching the renderer, and the updated tests in maintainer-recap-wire.test.ts exercise the real production path by asserting `report.totals.cohorts` is undefined and the formatted body omits cohort text end-to-end. The internal `totals.cohorts` shape is deliberately retained for callers that inject cohort-bearing reports directly, but `formatMaintainerRecap` no longer renders it under any circumstance, and the new redaction terms provide a second layer of protection if a cohort string ever reaches a summary line.

Nits — 5 non-blocking

• src/services/maintainer-recap.ts still computes and stores `totals.cohorts` (and `repo.cohorts`) internally even though no current call site ever populates it via `includeCohorts`, and `formatMaintainerRecap` never reads it — worth a one-line comment clarifying this is intentionally kept only for callers that inject reports directly, so a future reader doesn't assume it's dead code to delete.
• test/unit/maintainer-recap-format.test.ts's new cohort test manually injects `totals.cohorts` and a hand-built `summary` array with cohort strings rather than driving the whole thing through `buildMaintainerRecap`, so it verifies the formatter's redaction defense-in-depth but not that `buildMaintainerRecap` itself would ever produce such a summary line — the wire-level test compensates for this by covering the real end-to-end path.
• Consider a short code comment on `totals.cohorts`/`repo.cohorts` in src/services/maintainer-recap.ts explaining they're intentionally unrendered, to prevent a future cleanup pass from either deleting them or accidentally wiring them back into the formatter.
• If any other call site besides the scheduled job still passes `includeCohorts: true` to `loadGatePrecisionReport` (e.g. an internal/maintainer-only diagnostic endpoint), it'd strengthen the review to confirm that surface is genuinely maintainer-authenticated and never reachable from the public formatter path.
• src/services/maintainer-recap.ts:107: Keep the internal cohort aggregation only if another maintainer-authenticated caller still consumes `totals.cohorts`; otherwise consider removing that retained report shape in a follow-up to reduce private-field drift.

Why this is blocked

• 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.

CI checks failing

• validate
• validate-tests (1)

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

[cloudflare-workers-and-pages]

Deploying with    Cloudflare Workers

The latest updates on your project. Learn more about integrating Git with Workers.

Status	Name	Latest Commit	Preview URL	Updated (UTC)
✅ Deployment successful! View logs	gittensory-ui	7d14161	Commit Preview URL Branch Preview URL	Jul 12 2026, 09:57 AM