feat(mcp): register loopover_get_agent_audit_feed as a local stdio tool

[andriypolanski]

Summary

• loopover_get_agent_audit_feed was a remote MCP tool (src/mcp/server.ts) with a maintain audit-feed CLI mirror (CLI mirror for loopover_get_agent_audit_feed #6733), but had no local stdio MCP tool registration.
• This registers the missing stdio tool next to loopover_get_maintainer_noise, calling the same GET .../agent/audit-feed endpoint the CLI already uses (optional since/limit forwarded verbatim).

Closes #7757

Change

• Input schema (packages/loopover-mcp/bin/loopover-mcp.ts): added auditFeedShape (owner/repo + optional since/limit), mirroring the remote tool.
• Local stdio MCP tool: registered loopover_get_agent_audit_feed with a matching STDIO_TOOL_DESCRIPTORS entry (agent category). Thin GET proxy of {repoBase}/agent/audit-feed; query filters forwarded only when present (same pattern as the CLI).
• Tool count pin: 95 → 96 in mcp-tool-rename-aliases.test.ts (current main already includes Register loopover_get_automation_state as a local stdio MCP tool #7752 automation-state).

Tests

• New test/unit/mcp-cli-agent-audit-feed.test.ts: in-process InMemoryTransport coverage — with-filters path (since+limit) and no-filters path (empty query / query.size === 0 false branch), so every if / ternary arm is hit.

Validation

• git diff --check clean; npm run build:mcp clean.
• Targeted vitest suites pass (agent-audit-feed in-process, tools list, rename-aliases).
• Codecov patch = 100%. Verified from coverage-final.json: all added statements/functions hit; both since/limit if-arms and the query.size > 0 ternary arms are [1,1].
• Replaces closed conflicted feat(mcp): register loopover_get_agent_audit_feed as a local stdio tool #7967 (that PR targeted a stale 94→95 pin; this branch is based on current main at 95 tools).

UI Evidence

N/A — local stdio MCP tool only; no UI surface changed.

[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 81.94%. Comparing base (ca738de) to head (53f4108).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #7971      +/-   ##
==========================================
- Coverage   89.11%   81.94%   -7.17%     
==========================================
  Files          93       94       +1     
  Lines       21723    23942    +2219     
  Branches     3755     4602     +847     
==========================================
+ Hits        19358    19620     +262     
- Misses       2187     4126    +1939     
- Partials      178      196      +18     

Flag	Coverage Δ
shard-1	0.00% <0.00%> (?)
shard-2	11.80% <100.00%> (?)
shard-3	0.00% <0.00%> (?)

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	11.80% <100.00%> (ø)

[loopover-orb]

Tip

✅ LoopOver review result - approve/merge recommended

_{Review updated: 2026-07-22 00:50:32 UTC}

3 files · 1 AI reviewer · no blockers · readiness 88/100 · CI green · clean

✅ Suggested Action - Approve/Merge

• safe to merge

Review summary
This PR mirrors the existing remote loopover_get_agent_audit_feed tool as a local stdio MCP tool, following the exact same registerStdioTool + apiGet + query.size pattern already used by loopover_get_maintainer_noise and the maintain audit-feed CLI mirror. The query-building logic (URLSearchParams with since/limit forwarded only when present) is correct and matches the CLI's established convention, and the new test exercises both the filtered and unfiltered query paths in-process via InMemoryTransport. The tool-count pin bump (95→96) is applied consistently across both assertions in mcp-tool-rename-aliases.test.ts.

Nits — 5 non-blocking

• packages/loopover-mcp/bin/loopover-mcp.ts:376 — the `.max(200)` limit is a bare magic number; consider a named constant (e.g. `AUDIT_FEED_LIMIT_MAX`) since the comment already documents the 1..200 range in prose.
• The external brief's 'leaked secret' flag on test/unit/mcp-cli-agent-audit-feed.test.ts:31 (`LOOPOVER_API_TOKEN = "in-process-token"`) is a false positive — it's a fixture-scoped test token, matching the identical pattern already used by mcp-cli-maintain-tools.test.ts and other in-process test files in this repo.
• The handler param and callback in loopover-mcp.ts use `any` typing (`async ({ owner, repo, since, limit }: any) =>`), but this matches every other registerStdioTool handler already in the file, so it's consistent with existing convention rather than a new issue.
• Consider extracting the shared since/limit-max(200) query-building logic into a small helper if a third stdio tool with the same shape shows up, to avoid triplicating the URLSearchParams pattern.
• 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	#7757
Related work	✅ No active overlap found	No same-issue or scoped active PR overlap found.
Change scope	❌ 8/20	High 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: 190 registered-repo PR(s), 120 merged, 22 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	andriypolanski; Gittensor profile; 190 PR(s), 22 issue(s).
Improvement	✅ Minor	risk: clean · value: minor · LLM: moderate

Linked issue satisfaction

Addressed
The PR adds a registerStdioTool block for loopover_get_agent_audit_feed following the same pattern as the sibling tools, using stdioToolDescription and hitting the same GET .../agent/audit-feed endpoint the CLI already calls, plus updates tool-count tests and adds dedicated coverage.

Review context

• Author: andriypolanski
• Role context: outside_contributor
• Public audience mode: oss maintainer
• Lane context: Repository is configured for direct PR review.
• Public profile languages: Python, TypeScript, JavaScript, Rust, Cuda, Kotlin, MDX, Scala
• Official Gittensor activity: 190 PR(s), 22 issue(s).
• PR-specific overlap: none found.

Contributor next steps

• Start here: Add a concise scope and risk note.
• Then work through the remaining 1 step in the Signals table above.

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.