feat(mcp): mirror loopover_monitor_open_prs to the stdio wrapper and CLI

[reyanthony062001-ops]

Closes #6732. Supersedes #6781, which was auto-closed on a CI failure that came from its base, not its diff — see the note at the end.

loopover_monitor_open_prs (src/mcp/server.ts) and GET /v1/contributors/:login/open-pr-monitor (src/api/routes.ts, gated by requireContributorAccess) already do this work, but the stdio wrapper registered no matching tool and the CLI had no command — so a contributor could not review their own open PRs from the terminal. This mirrors the existing surface; it adds no route, no dependency, and no source upload.

What changed

• registerStdioTool("loopover_monitor_open_prs", ...) in packages/loopover-mcp/bin/loopover-mcp.js, proxying the existing route via apiGet and reusing loginShape. Its descriptor carries the same description the MCP tool already publishes, under category: "discovery" alongside loopover_get_decision_pack. Tool count 63 → 64.
• loopover-mcp monitor-open-prs --login <github-login> [--json], resolving the login from --login / LOOPOVER_LOGIN / GITHUB_LOGIN exactly as decisionPackCli does, registered in CLI_COMMAND_SPEC (so shell completion and suggestCommand's "did you mean" pick it up) and listed in --help + the Environment block.
• Summary parity: the CLI prints the API's own summary rather than composing a second sentence, so the two mirrored surfaces cannot drift apart for one payload. A test pins that.
• #6261 sanitization: the plain-text path passes API-chosen text through sanitizePlainTextTerminalOutput — the API composes summary/guidance and echoes PR titles back from third-party repos, so a hostile string must not be able to repaint the terminal. --json deliberately stays raw, since JSON.stringify escapes U+001B and callers parse that contract.

Tests

New test/unit/mcp-cli-monitor-open-prs.test.ts (8 tests, in the mcp-cli-* style): tool registration, route+method proxied, output parity between the mirrored surfaces for identical input, plain-text rendering, login resolution via both env vars, the shared login-required failure, ANSI sanitization (plain-text stripped / --json preserved), and --help/completion documentation. The fixture route + openPrMonitorFixture() mirror the real ContributorOpenPrMonitor shape.

Each new test was verified to fail against the un-mirrored bin before passing, and the sanitization test was verified to fail against an un-sanitized build — they catch the defects they describe rather than merely passing.

Validation

Rebased onto ab84f78 — the first green main after #6785 fixed the selfhost-grafana-dashboard assertions.

• Full npx vitest run backend suite on this base: green, including test/unit/selfhost-grafana-dashboard.test.ts (30/30), the suite that failed on feat(mcp): mirror loopover_monitor_open_prs to the stdio wrapper and CLI #6781's base
• npm run typecheck: clean · npm run build:mcp + npm run test:mcp-pack: pass
• loopover-mcp tools --json reports 64, matching the bumped assertion in mcp-tool-rename-aliases.test.ts

Why #6781 was closed, and what changed here

#6781 carried this same diff. Its validate-tests (6) shard failed on test/unit/selfhost-grafana-dashboard.test.ts — six assertions in a file this diff does not touch. That break arrived with #6779 (b440c84) at 05:29Z and left main's own ci.yml red (42cca18, 67af510); #6781 was branched from 67af510 and inherited it. #6785 has since fixed those assertions and main is green again at ab84f78, which is this PR's base — so the shard-6 failure is gone rather than argued away.

The only change from #6781 is the base. The diff is byte-identical.

Note on the issue text

The issue specifies the assertion bump as 63 → 64, which is exactly right; this PR does that. It also asks for 99% patch coverage on "the new route/tool/CLI code" — worth flagging that the route and MCP tool already exist and are unchanged here, and packages/loopover-mcp/bin/** is outside the coverage include in vitest.config.ts, so this diff adds no Codecov-measured lines (codecov/patch passed on #6781 for that reason). The behaviour is covered by the unit tests above regardless.

[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 93.63%. Comparing base (ab84f78) to head (77f3122).
⚠️ Report is 3 commits behind head on main.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #6789   +/-   ##
=======================================
  Coverage   93.63%   93.63%           
=======================================
  Files         683      683           
  Lines       68148    68148           
  Branches    18703    18703           
=======================================
  Hits        63812    63812           
  Misses       3350     3350           
  Partials      986      986           

Flag	Coverage Δ
shard-1	43.66% <ø> (ø)
shard-2	36.83% <ø> (ø)
shard-3	32.74% <ø> (ø)
shard-4	34.67% <ø> (ø)
shard-5	31.00% <ø> (ø)
shard-6	45.67% <ø> (ø)

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

[loopover-orb]

Tip

✅ LoopOver review result - approve/merge recommended

_{Review updated: 2026-07-17 06:42:58 UTC}

6 files · 1 AI reviewer · no blockers · readiness 82/100 · CI green · clean

✅ Suggested Action - Approve/Merge

• safe to merge

Review summary
This PR mirrors the existing loopover_monitor_open_prs API/MCP-tool surface into the stdio wrapper and CLI, following the established pattern used for decision-pack/repo-decision: tool registration, CLI command spec, login resolution via --login/LOOPOVER_LOGIN/GITHUB_LOGIN, ANSI-sanitized plain-text output vs raw --json, and a summary-parity contract so the CLI doesn't compose a second sentence. The diff is well-tested (8 new tests covering registration, route proxying, mirror parity, ANSI stripping, and help/completion text) and consistent with prior PRs in this series (tool count 63→64 bumped correctly across all three counting tests). Code is correctly wired: getOpenPrMonitor hits the right route, openPrMonitorToolSummary is shared between the stdio tool and CLI, and sanitizePlainTextTerminalOutput is applied to every API-chosen string on the plain-text path while --json stays raw.

Nits — 5 non-blocking

• packages/loopover-mcp/bin/loopover-mcp.js: monitorOpenPrsCli's nested loops (guidance, pullRequests, nextSteps) push its cyclomatic complexity to ~12 per the external brief; consider extracting a small `printOpenPr(pr)` helper for readability, though it's not a correctness issue.
• The PR description says 'Closes CLI mirror for loopover_monitor_open_prs #6732' but doesn't show the actual issue text — worth double-checking the CLI flag name/shape (--login, --json) matches whatever the issue requested.
• test/unit/mcp-cli-monitor-open-prs.test.ts: the ANSI-stripping test asserts `plain).not.toContain("")` which is a slightly fragile way to check for the absence of the escape character — an explicit regex like `/\x1b/` would be clearer intent, though this is purely stylistic.
• Consider factoring the repeated per-PR printing block in monitorOpenPrsCli into a helper to reduce the flagged complexity and ease future changes to the output format.
• If not already covered elsewhere, add a brief note in README or --help about what 'cleanupFirst'/'pendingScenarios' fields mean for --json consumers, since they're part of the payload contract now surfaced via CLI.

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	#6732
Related work	⚠️ 3 scoped overlaps	Top overlaps are listed below; lower-confidence bulk is hidden.
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: 37 registered-repo PR(s), 23 merged, 1 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	reyanthony062001-ops; Gittensor profile; 37 PR(s), 1 issue(s).
Improvement	✅ Minor	risk: clean · value: minor · LLM: moderate

Linked issue satisfaction

Addressed
The PR adds registerStdioTool("loopover_monitor_open_prs", ...) proxying the existing GET /v1/contributors/:login/open-pr-monitor route, adds a monitor-open-prs CLI command resolving login from --login/LOOPOVER_LOGIN/GITHUB_LOGIN matching decisionPackCli, supports --json, updates the tool-count assertion from 63 to 64, and includes a new mcp-cli-monitor-open-prs.test.ts with an explicit output-par

Review context

• Author: reyanthony062001-ops
• Role context: outside_contributor
• Public audience mode: oss maintainer
• Lane context: Repository is configured for direct PR review.
• Public profile languages: Cuda, HTML, JavaScript, Python, TypeScript
• Official Gittensor activity: 37 PR(s), 1 issue(s).
• Related work: Titles/paths share 7 meaningful terms. (issue #6732, issue #6747)
• Related work: Titles/paths share 7 meaningful terms. (issue #6732, issue #6741)
• Related work: Titles/paths share 8 meaningful terms. (issue #6732, issue #6751)
• Additional title-only matches omitted; title-only overlap does not block.

Contributor next steps

• Start here: Review top overlaps.
• Then work through the remaining 2 steps 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.