菜单
Runs detect_subagents then run_subagent (codex/gemini) in review mode via subagent-orchestrator MCP, reads .agent-runs outputs, and summarizes findings. Use when the user asks for external CLI review, codex/gemini review, subagent panel, tie-breaker, or invokes /check-subagents, /run-codex-review, /run-gemini-review, /run-subagent-review.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Bump semver + promote CHANGELOG Unreleased via compiled harness. Use for /release, version bumps, or tagging. Runs scripts/agent_skills/release_bump.py; LLM only when Unreleased is ambiguous. Never commit/push unless the user asks.
Vexlum Scoring MCP — compact search+dispatch on is-be-mcp / is-be-live and gallery is-ui-mcp / is-ui-live.
Choose CLI tools and safe patterns before repo work. Install checklist, bounded search/read commands, and confirmation gates. Use when navigating the codebase, running shell commands, or setting up agent tooling on Windows, WSL2, or macOS.
Inspect JSON, YAML, and API responses with jq, yq, dasel, curl, and httpie. Use for package.json, config.json, and backend health checks. Never output secrets or full credential-bearing config.
Backend build, test, lint, and verify workflows via WSL pytest, ruff, doctor, and docker compose. Use before claiming work complete or opening a PR. Do not invent ad hoc Python commands outside documented venvs.
Choose and run the right search tool: rg vs grep vs ast-grep vs fd vs fzf. Bounded repo navigation with ripgrep, fd, tree, bat, and structural search. Use when finding symbols, files, syntax patterns, or deciding which CLI search to use before reading or editing code.
| name | subagent-review |
| description | Runs detect_subagents then run_subagent (codex/gemini) in review mode via subagent-orchestrator MCP, reads .agent-runs outputs, and summarizes findings. Use when the user asks for external CLI review, codex/gemini review, subagent panel, tie-breaker, or invokes /check-subagents, /run-codex-review, /run-gemini-review, /run-subagent-review. |
Run external CLI coding agents through the subagent-orchestrator MCP server — not by spawning codex / gemini / claude directly in the terminal.
../subagent-orchestrator/agent-orchestrator built (npm install && npm run build there once).cursor/mcp.json as cli-review (or user-level subagent-orchestrator) and reloaded in CursorBefore CallMcpTool, find the enabled server where serverName is subagent-orchestrator (project key cli-review; user-level keys may also appear). Read tool schemas from mcps/*/tools/detect_subagents.json and run_subagent.json. Prefer the project server so WORKSPACE_ROOT is this repo.
Call detect_subagents (no arguments). Present a short table:
| Agent | Available | Mode | Notes |
|---|
Note: Claude is detection-only in v0.1 — do not rely on it for live run_subagent.
@ mentions (max 20; no .env, keys, binaries)review (default) or tie-breaker when comparing opinionsfalsetrue when user says “dry run” or you are validating setupCall run_subagent with JSON per references/mcp-tools.md.
For a “panel” (user wants both agents): run sequentially — codex then gemini (or vice versa) with the same task and files. There is no run_subagent_panel tool in v0.1.
outputFile, summary, ok, error)..agent-runs/.../stdout.md at the returned path.## External review summary
### Verdict
### Findings (blocker | high | medium | low | nit)
### Suggested changes
### Agent agreement (if multiple runs)
### Next steps
Do not apply patches unless the user explicitly asks.
git diff --name-only under modules/, tests/, migrations/, frontend/src/ when the user did not @ paths.| Situation | Action |
|---|---|
| Agent not on PATH | Report detect_subagents notes; suggest install/auth |
allowWrites rejected | Explain v0.1 is review-only |
| Claude run requested | Use Codex or Gemini instead |
| MCP server missing | docs/technical/EXTERNAL_CLI_REVIEWS.md; build sibling orchestrator |
external-cli-subagents.mdc/check-subagents, /run-codex-review, /run-gemini-review, /run-subagent-review../subagent-orchestrator/agent-orchestrator/README.md