Cognee MCP (Model Context Protocol) integration provides seamless connectivity between Cognee’s knowledge graph platform and MCP-compatible AI development tools like Claude, Cursor, and Continue.
MCP enables AI assistants to securely access and manage your knowledge graphs, creating persistent memory across conversations.

​
Quick Start

Get up and running with Cognee MCP in minutes using Docker.

​
Prerequisites

Before you begin, ensure you have:

Docker

Docker installed and running on your system

OpenAI API Key

Valid OpenAI API key for LLM operations

​
Environment Setup

First, set your OpenAI API key: 
export OPENAI_API_KEY=your_api_key_here
Replace your_api_key_here with your actual OpenAI API key. Keep this secure and never commit it to version control.

​
Launch the Server

Start the Cognee MCP server with a single command: 
docker run -e TRANSPORT_MODE=http --env-file ./.env -p 8000:8000 --rm -it cognee/cognee-mcp:main
The server will run on port 8000. Deleting the container will remove all stored memory data.

​
How Cognee MCP Works

Built around the Model Context Protocol (MCP), the Cognee MCP Server exposes a standardized set of memory management tools that any MCP-compatible AI assistant can use.

​
Available Tools

Memory Management

​
Setup Options

Choose your preferred deployment method:

Docker (Quick Start)

Recommended for most usersReady-to-use container with all dependencies included. Perfect for getting started quickly.

Local Setup

For developers & customizationBuild from source for development, custom configurations, and latest features.

​
Next Steps

Local MCP Setup

Deploy from source with full control and latest features

IDE Integrations

Connect with Cursor, Claude, Continue, and more