Skip to content
Get packs

Managing MCP Servers

Add, edit, and troubleshoot MCP tool servers in ThinkRail's config files

Source Download skill
Repository
JetBrains/thinkrail
Stars
7,089
Licence
Apache-2.0
Files
1 file

What it does

This skill teaches the agent how to set up and manage MCP servers, which are add-ons that give an AI assistant extra tools such as documentation lookups, browser control, or GitHub access. The agent edits the JSON configuration files (a user-level file or a repository-level .pi/mcp.json) instead of running commands, and sends the user to the Settings screen for approvals and sign-ins. It also covers keeping secrets out of files, pinning package versions, and explaining how MCP tools are called.

Great for

  • ThinkRail users who want to connect a new MCP server such as GitHub or a documentation source
  • People whose MCP tool is missing, needs sign-in, fails, or was blocked and want the agent to diagnose it
  • Teams that want to share a server configuration through a repository file

Skip it if

  • Users of other AI tools that do not use ThinkRail or the pi MCP format; the paths and approval flow are specific to ThinkRail, so the instructions will not match their setup
  • Anyone who would paste real API tokens into the chat expecting the agent to store them safely; the skill says to use environment variables or sign-in instead, and the agent should not be trusted with raw secrets
  • Repository servers from untrusted projects: a server in .pi/mcp.json runs only after the user approves it, so approval should be given only for servers the user has reviewed

Try asking

  • “Add the GitHub MCP server to my ThinkRail setup using my GitHub CLI login”
  • “My docs MCP tool is missing in this chat, can you check the configuration?”
  • “Set up a Playwright browser MCP server for my user account with a pinned version”
  • “Share this project's MCP server with my team through the repository config”
Step-by-step guide: how to use Managing MCP ServersInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the Apache-2.0 licence and an attribution note.

    managing-mcp-servers.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip managing-mcp-servers.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip managing-mcp-servers.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Ask for read-only access first; the GitHub recipe defaults to the read-only endpoint and write-capable tools need explicit user request.
  • Keep personal servers in the user-level file and only use the repository file when the whole team should share the server.
  • If the GitHub server rejects the GitHub CLI token, set GITHUB_TOKEN in your environment rather than pasting the token into a file.

What's inside

  • SKILL.md

Redistributed unmodified from JetBrains/thinkrail under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-11-02. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.