Start free

Add the Cognee MCP server to Cursor

Config last verified Nov 10, 2026

The exact config to run Cognee in Cursor — paste it in, restart, and the tools load.

Prerequisites

  • Cursor installed.
  • LLM_API_KEY — API key for the LLM provider Cognee uses to extract the graph and answer queries (OpenAI by default).

Setup

1. Open ~/.cursor/mcp.json

2. Add this configuration

Add to ~/.cursor/mcp.json

~/.cursor/mcp.json
json
{
  "mcpServers": {
    "cognee": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "TRANSPORT_MODE=stdio",
        "-e",
        "LLM_API_KEY",
        "cognee/cognee-mcp:main"
      ],
      "env": {
        "LLM_API_KEY": "<LLM_API_KEY>"
      }
    }
  }
}

3. Restart Cursor and confirm the Cognee tools load.

Gotchas

Cursor keeps MCP servers under an "mcpServers" object in ~/.cursor/mcp.json for global use, and also reads a project-scoped .cursor/mcp.json so a repository can ship its own servers. Remote servers are configured with a bare "url" field and connect natively, with OAuth handled in-app when a server requires it.

← Back to the Cognee MCP server
MCP Server DirectoryGlen home