cognee-mcp

Level ALocal

Cognee MCP server

Overview

cognee-mcp runs cognee's memory engine as a Model Context Protocol server, letting MCP-capable terminal or IDE clients build and query memory for AI agents. It exposes a minimal memory API: remember stores session cache or permanent graph memory, recall retrieves it, and forget removes a dataset or all owned memory. A code_search tool queries indexed symbols, dependencies, architecture, and change impact. It supports stdio, SSE, and Streamable HTTP transports, and can connect to Cognee Cloud or…

⭐ 33kDevelopment ToolsPythonRepo ↗Updated 2026-11-02

Installation

MCP client config (paste into mcp.json)

{
  "mcpServers": {
    "cognee": {
      "command": "uvx",
      "args": [
        "cognee-mcp@latest"
      ]
    }
  }
}

Agent install prompt

Copy and paste into Cursor / Claude / other agents to install this MCP for you

Please read this MCP repository and install it into my current environment:
https://github.com/topoteretes/cognee/tree/main/cognee-mcp

Known working client config (can write into mcp.json):
{
  "mcpServers": {
    "cognee": {
      "command": "uvx",
      "args": [
        "cognee-mcp@latest"
      ]
    }
  }
}

Please: 1) check README for deps and env vars; 2) write config for my client; 3) if secrets are missing, tell me what to fill — do not invent values.

Related servers