home / skills / laurigates / claude-plugins / mcp-management

This skill helps you install and configure MCP servers for Claude Code projects, enabling connections, updates, and troubleshooting per project.

npx playbooks add skill laurigates/claude-plugins --skill mcp-management

Review the files below or copy the command above to add this skill to your agents.

Files (1)
SKILL.md
681 B
---
model: haiku
created: 2025-12-16
modified: 2026-02-05
reviewed: 2025-12-16
name: mcp-management
description: |
  Install and configure Model Context Protocol (MCP) servers for Claude Code projects.
  Use when you want to add or enable an MCP server, connect a tool or integration
  (database, API, file system), update MCP settings in .mcp.json, or troubleshoot
  MCP server connection issues.
allowed-tools: Bash, Read, Write, Edit, Grep, Glob, AskUserQuestion
---

# MCP Server Management

Expert knowledge for managing Model Context Protocol (MCP) servers on a project-by-project basis, with intelligent suggestions based on project context.

[Rest of content continues...]

Overview

This skill installs and configures Model Context Protocol (MCP) servers for Claude Code projects, enabling reliable context sharing and tool integrations. It helps add or enable MCP servers, wire up databases/APIs/filesystems, update .mcp.json settings, and diagnose connection issues quickly. The focus is practical: get an MCP server running, connected, and validated in your project environment.

How this skill works

The skill inspects project files and the .mcp.json manifest to determine required MCP endpoints, credentials, and integrations. It provides commands and configuration edits to install or enable an MCP server, create or update integration blocks (database, API, file system), and apply environment-specific settings. For troubleshooting, it runs connectivity checks, validates auth tokens, and surfaces actionable error messages and recommended fixes.

When to use it

  • Adding a new MCP server to a Claude Code project
  • Connecting a database, API, or filesystem to an MCP server
  • Updating or versioning MCP settings in .mcp.json
  • Enabling MCP access for a CI/CD or local dev environment
  • Diagnosing failed MCP connections or auth problems

Best practices

  • Keep .mcp.json under version control and use clear environment overlays (dev/staging/prod).
  • Store credentials in a secure secrets manager and reference them via environment variables—avoid committing secrets.
  • Validate MCP endpoint and auth with automated checks after changes, and include them in CI tests.
  • Use minimal required scopes for service accounts and rotate credentials regularly.
  • Document any nonstandard integrations and required database schemas or API contracts in the project docs.

Example use cases

  • Install an MCP server locally for development and link it to a test PostgreSQL instance.
  • Enable MCP in a staging environment and connect an external analytics API for context enrichment.
  • Update .mcp.json to change context retention or cache settings and apply the change across environments.
  • Troubleshoot a failing MCP connection by checking TLS, endpoint reachability, and token expiration.
  • Migrate MCP settings from one host to another and ensure integrations (DB, files) are reattached cleanly.

FAQ

Where are MCP settings stored?

.mcp.json in the project root holds MCP endpoints, integrations, and environment overlays. Keep it in Git but exclude secrets; reference secrets via env variables or a secrets manager.

How can I test an MCP connection after configuration?

Use the provided connectivity checks to ping the MCP endpoint, validate TLS certificate, and perform an authenticated test request. Include smoke tests in CI to catch regressions.