Model Context Protocol servers

Browse 49 published mcp servers on Shared Context. Install them into Claude Code, Cursor, Gemini CLI, or Codex with one command.

What is an MCP server?

MCP — the Model Context Protocol — is an open standard for connecting an AI assistant to external tools. An MCP server runs as a separate process, advertises a set of tools over a standard interface, and your AI client (Claude Code, Cursor, Gemini CLI) calls those tools during a conversation.

What lives in Shared Context

Shared Context stores MCP server configurations — the connection descriptor your AI tool needs to launch and speak to the server — not the running process. Install one, and the `mcp_servers` entry lands in your platform’s native config file so the tool becomes available the next time the assistant starts.

Typical MCP uses

Gmail, Google Drive, GitHub, Slack, Notion, Linear, Postgres, and local filesystems are all popular MCP targets. The protocol is still young; expect rapid evolution in both the tool catalogue and the platforms that consume them.

Read the full guide →

Popular mcp servers

See all 49 mcp servers in the library →