feat(mcp): add remote + stdio + CLI surfaces for the selftune override audit

[michiot05]

Summary

• Closes Self-tune override audit trail (selftune/overrides/audit) has no remote MCP tool, local stdio MCP tool, or CLI verb #7798: give the self-tune override audit trail the same three-surface reach as its gate-precision sibling — remote MCP, local stdio MCP, and a maintain selftune-audit CLI verb — instead of raw REST only.
• Remote tool (src/mcp/server.ts): loopover_get_selftune_override_audit, category "maintainer", input owner/repo + optional limit (mirroring the route's ?limit query), calling the same listOverrideAudit the REST route uses and returning its exact { repoFullName, audit } shape. Same per-repo read gate as loopover_get_gate_precision (requireRepoAccess), per the issue's required pattern.
• Stdio tool (packages/loopover-mcp/bin/loopover-mcp.ts): proxies GET {repoBase}/selftune/overrides/audit via apiGet, forwarding ?limit only when set — same shape as the loopover_get_gate_precision stdio registration.
• CLI verb: maintain selftune-audit --repo owner/repo [--limit N] [--json], following the precision subcommand's structure exactly (same emit/--json handling, same query-building), wired into the subcommand list, maintain --help, the unknown-subcommand message, and the top-level usage line.
• listOverrideAudit, the REST route, and its requireRepoMaintainer gate are untouched — this only adds the three missing mirror surfaces.
• Tests: a remote-tool regression test mirroring loopover_get_gate_precision's own (test/unit/mcp-selftune-override-audit.test.ts: repo-scoped rows + limit pass-through, empty trail, forbidden identity), plus in-process bin coverage for the CLI dispatcher AND stdio tool via the exported maintainCli / server + InMemoryTransport (test/unit/mcp-cli-selftune-audit.test.ts, same loopover_plan_repo_issues has no REST route, CLI command, or local stdio MCP tool #7764 entrypoint-guard pattern as mcp-cli-plan-issues), covering both sides of every new branch (--limit present/absent, detail-bearing/detail-less events, audit-less payload, unknown-subcommand fall-through, help text). Tool-count pin bumped 100 → 101 in mcp-tool-rename-aliases; packages/loopover-mcp/bin/** confirmed in vitest coverage.include.

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 a currently open issue this PR resolves (e.g. Closes #123) — a linked open issue is required for every contributor PR.

Validation

• git diff --check
• npm run actionlint
• npm run typecheck
• npm run test:coverage locally; codecov/patch requires ≥99% coverage of the lines AND branches you changed (aim for 100% on your diff so CI variance does not fail near the threshold). Global coverage is a non-blocking trend with a loose 90% backstop, not the gate.
• npm run test:workers
• npm run build:mcp
• npm run test:mcp-pack
• npm run ui:openapi:check
• npm run ui:lint
• npm run ui:typecheck
• 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, explain why:

• None skipped — the full npm run test:ci gate was run from the repo root (it includes every check above plus the drift checks) and passed, followed by npm audit --audit-level=moderate. The audit reports pre-existing advisories already present on main's dependency tree (fast-uri, and @hono/node-server transitively under @modelcontextprotocol/sdk / @cloudflare/codemode / agents) — this PR changes no dependencies, and those advisories are already being addressed by the open Renovate updates.

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 — backend MCP/CLI surfaces only; no visible UI, frontend, docs, or extension change.

Notes

• MCP contract note: tools/list gains loopover_get_selftune_override_audit (category maintainer) with an output schema (repoFullName, audit); audit rows stay schema-opaque (z.unknown()) because listOverrideAudit is their single source of truth, matching gatePrecisionOutputSchema's sub-report pattern. No REST or OpenAPI surface changed, so no regenerated artifacts were needed (command-reference:check, docs:drift-check, manifest:drift-check all pass unchanged).
• The forbidden path is an error result (same as the gate-precision family's requireRepoAccess throw), so the output schema carries no status arm.

[superagent-security]

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

[loopover-orb]

Tip

✅ LoopOver review result - approve/merge recommended

_{Review updated: 2026-07-22 03:34:40 UTC}

9 files · 1 AI reviewer · no blockers · readiness 100/100 · CI green · clean

✅ Suggested Action - Approve/Merge

• safe to merge

Review summary
This PR mirrors an already-live REST route (`GET /selftune/overrides/audit`) onto the remote MCP tool, stdio MCP tool, and CLI verb, closely following the existing `loopover_get_gate_precision`/`precision` triad pattern end-to-end (shape, category, gate, help text, tool-count pin). The wiring is consistent across all three surfaces — same `requireRepoAccess` gate on the remote tool, same `?limit` forwarding logic on stdio/CLI, and the tool-count assertions (100→101) and CLI usage/help strings are updated everywhere they appear. Tests cover both branches of the new `--limit` logic, detail-bearing vs detail-less audit rows, the audit-less payload fallback, and the unknown-subcommand fall-through.

Nits — 4 non-blocking

• test/unit/mcp-cli-selftune-audit.test.ts:35 sets `LOOPOVER_API_TOKEN = "in-process-token"` for the fixture harness — this is a hardcoded test placeholder, not a real credential, but worth confirming it matches the existing convention in sibling `mcp-cli-*.test.ts` files.
• packages/loopover-mcp/bin/loopover-mcp.ts:3148 and :3958 use `any` for the stdio tool handler and CLI event-mapping params — consistent with the file's existing style for sibling tools (e.g. gate-precision), so not a new deviation, but still loose typing worth tightening file-wide at some point.
• Consider narrowing the `any` types on the new `registerStdioTool` handler (packages/loopover-mcp/bin/loopover-mcp.ts:3148) and the `audit.map((event: any) => ...)` callback (packages/loopover-mcp/bin/loopover-mcp.ts:3958) to a shared `OverrideAuditEvent` type instead of `any`, matching the return shape from `listOverrideAudit`.
• Possible secret-shaped assignment in the diff (generic_secret_assignment) — Verify the value is not a real credential.

Decision drivers

• ✅ Code review — No blockers (1 reviewer)
• ✅ Gate result — Passing (No configured blocker found.)

Context & advisory signals — never blocks the verdict

Signal	Result	Evidence
Linked issue	✅ Linked	#7798
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 (1 linked issue).
Validation posture	✅ 25/25	PR body includes validation/test evidence.
Contributor workload	✅ 10/10	Author activity: 8 registered-repo PR(s), 1 merged, 2 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	michiot05; Gittensor profile; 8 PR(s), 2 issue(s).
Improvement	✅ Minor	risk: clean · value: minor · LLM: moderate

Linked issue satisfaction

Partially addressed
The PR adds the remote MCP tool, stdio MCP tool, and CLI verb for the override audit, closely mirroring gate-precision's three surfaces as requested, but it deviates from the issue's explicit gate requirement by using requireRepoAccess instead of the same requireRepoMaintainer/requireRepoAccess gate the issue says must mirror loopover_get_gate_precision exactly (the issue emphasizes the same requi

Review context

• Author: michiot05
• Role context: outside_contributor
• Public audience mode: oss maintainer
• Lane context: Repository is configured for direct PR review.
• Public profile languages: not available
• Official Gittensor activity: 8 PR(s), 2 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.
• 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.

🧪 Chat with LoopOver

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

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

Full command reference: https://loopover.ai/docs/loopover-commands

🧪 Experimental — new and may change.

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

• Re-run LoopOver review

[loopover-orb]

LoopOver approves — the gate is satisfied and CI is green.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 83.13%. Comparing base (7c1a21e) to head (67f30bf).
⚠️ Report is 2 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #7997      +/-   ##
==========================================
- Coverage   91.91%   83.13%   -8.78%     
==========================================
  Files         739       96     -643     
  Lines       75854    25176   -50678     
  Branches    23018     4802   -18216     
==========================================
- Hits        69723    20931   -48792     
+ Misses       5038     3994    -1044     
+ Partials     1093      251     -842     

Flag	Coverage Δ
shard-1	23.50% <22.72%> (-34.95%)	⬇️
shard-2	23.11% <45.45%> (-28.91%)	⬇️
shard-3	35.67% <77.27%> (-14.37%)	⬇️

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

Files with missing lines	Coverage Δ
packages/loopover-mcp/bin/loopover-mcp.ts	17.96% <100.00%> (+0.51%)	⬆️
src/mcp/server.ts	96.73% <100.00%> (+0.02%)	⬆️

... and 643 files with indirect coverage changes