Claude Code MCP Servers
Browse Model Context Protocol servers that connect Claude Code to databases, APIs, files, and external tools.
MCP servers connect Claude Code to external systems via the Model Context Protocol. Configure them in .mcp.json (project-scoped, committed to git) or ~/.claude.json (user-scoped). Once configured, their tools appear as callable functions in your Claude session.
No MCP servers match your search. Try a different query or filter.
Getting Started
Project-scoped: .mcp.json in the project root. User-scoped: ~/.claude.json.
{ "mcpServers": { "<name>": { "command": "...", "args": [...] } } }
Restart the session (or run /mcp) to load the server and inspect its connection status.
Frequently Asked Questions
What are Claude Code MCP servers?
MCP (Model Context Protocol) servers are external processes that expose tools, resources, and prompts to Claude. They let Claude read databases, query APIs, and operate external systems safely through a standardized protocol.
How do I install an MCP server?
Add an entry to .mcp.json in your project root (or ~/.claude.json for global use). Each entry specifies a command, args, and optional env. Restart Claude Code to load the server.
What is the difference between project and user MCP servers?
Project servers (.mcp.json) are committed to git and shared with your team. User servers (~/.claude.json) are personal and available across all your projects.
Are MCP servers safe?
MCP servers run as local processes with the permissions of your user. Review the server's source and license before installing. The security badges on each card summarize our automated checks.