mcp-maintainer — Claude Code skill: MCP server maintenance
The mcp-maintainer skill guides maintenance of the repository’s local MCP server, from checking its API to validating tool changes. Use it when updating MCP tools or investigating runtime issues.
MIT ★ 1,198 updated 2026-10-23
npx skills add hoangsonww/Claude-Code-Agent-Monitor --skill mcp-maintainer -a claude-codeWhat mcp-maintainer does
The skill lays out a workflow for the local MCP server. Check the dashboard API at /api/health, then inspect the relevant modules under mcp/src/tools/domains/. Changes to safety checks belong alongside the existing guards in mcp/src/policy/tool-guards.ts.
Run npm run mcp:typecheck and npm run mcp:build to validate changes. The skill also points to a tool-domain map and an operations runbook: references/tool-domain-map.md and references/operations-runbook.md.
Keep target checks restricted to loopback, and require explicit flags for tools that mutate or destroy data. Protocol data must stay out of stdout. Removing that boundary can expose data through the server’s output.
What it helps with
- check whether the MCP dashboard API responds
- locate the domain module behind a tool
- preserve guards while changing policy checks
- type-check and build the MCP server
Example prompts for the agent
- Check the dashboard API at
/api/healthand locate the affected tool module. - Update the guard in
mcp/src/policy/tool-guards.tswhile retaining loopback-only target checks. - Run
npm run mcp:typecheckandnpm run mcp:buildafter the changes.
How to install mcp-maintainer
1Skills CLI (Vercel)
npx skills add hoangsonww/Claude-Code-Agent-Monitor --skill mcp-maintainer -a claude-code2Manually: put the files in place
~/.claude/skills/mcp-maintainer/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/mcp-maintainer/
1Skills CLI (Vercel)
npx skills add hoangsonww/Claude-Code-Agent-Monitor --skill mcp-maintainer -a codex2Manually: put the files in place
.agents/skills/mcp-maintainer/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add hoangsonww/Claude-Code-Agent-Monitor --skill mcp-maintainer -a cursor2Manually: put the files in place
.cursor/skills/mcp-maintainer/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install hoangsonww/Claude-Code-Agent-Monitor .agents/skills/mcp-maintainer2Manually: put the files in place
.github/skills/mcp-maintainer/SKILL.md1Skills CLI (Vercel)
npx skills add hoangsonww/Claude-Code-Agent-Monitor --skill mcp-maintainer -a gemini-cli2Platform CLI
gemini skills install https://github.com/hoangsonww/Claude-Code-Agent-Monitor --path .agents/skills/mcp-maintainerManually: put the files in place
.devin/skills/mcp-maintainer/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about mcp-maintainer
How do I install mcp-maintainer in Claude Code?
Run npx skills add hoangsonww/Claude-Code-Agent-Monitor --skill mcp-maintainer -a claude-code. The skill file is at ~/.claude/skills/mcp-maintainer/SKILL.md.
What does mcp-maintainer do?
It provides a workflow for maintaining this repository’s local MCP server: checking the dashboard API, inspecting tool modules and validating changes.
What safety checks does the skill cover?
Target checks must remain loopback-only, while mutating and destructive tools require explicit flags. Protocol data must not be written to stdout.
Source text
SKILL.md · MIT
--- name: mcp-maintainer description: Operate and maintain the local MCP server for this repository. Use for MCP tool updates, policy-guard changes, host configuration, and MCP runtime troubleshooting. --- # MCP Maintainer Skill ## Workflow - Confirm dashboard API availability (`/api/health`). - Inspect affected MCP domain modules under `mcp/src/tools/domains/`. - Preserve safety gates in `mcp/src/policy/tool-guards.ts`. - Validate with `npm run mcp:typecheck` and `npm run mcp:build`. ## Safety rules - Keep loopback-only target checks enabled. - Keep mutating and destructive tools behind explicit flags. - Do not log protocol data to stdout. ## References - `references/tool-domain-map.md` - `references/operations-runbook.md`
© hoangsonww · MIT · hoangsonww/Claude-Code-Agent-Monitor