Graphiti
MCP server for graphiti
What is Graphiti?
Graphiti is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to mcp server for graphiti
MCP server for graphiti
This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- MCP server for graphiti
Use Cases
Maintainer
Works with
Installation
NPM
npx -y graphitiManual Installation
npx -y graphitiConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Graphiti
Graphiti MCP Server provides a dynamic knowledge graph backend powered by Neo4j and OpenAI embeddings, exposed through the Model Context Protocol. It lets AI assistants store facts, entities, and relationships as graph nodes, then query them semantically to recall context across long sessions or multiple agents. This is particularly valuable for coding assistants and research agents that need to remember decisions, architecture choices, or factual information beyond a single conversation window.
Prerequisites
- Docker and Docker Compose installed (Neo4j and the server run as containers)
- An OpenAI API key (used for embeddings and the LLM backbone; gpt-4.1-mini by default)
- Git to clone the repository
- An MCP-compatible client such as Claude Desktop, Cursor, or Claude Code
Clone the repository
Clone the graphiti-mcp-server repository to your local machine. The project ships a Docker Compose file that starts both Neo4j and the MCP server.
git clone https://github.com/gifflet/graphiti-mcp-server.git
cd graphiti-mcp-serverConfigure environment variables
Copy the sample environment file and fill in your OpenAI API key. All other defaults (Neo4j credentials, model names) work out of the box for local development.
cp .env.sample .env
# Edit .env and set:
# OPENAI_API_KEY=sk-...
# NEO4J_PASSWORD=demodemo (or your preferred password)Start the services with Docker Compose
Spin up Neo4j and the Graphiti MCP server together. The server will be available at http://localhost:8000 via SSE transport.
docker compose up -dVerify the server is healthy
Check that the server started correctly by hitting the health endpoint.
curl http://localhost:8000/healthConfigure your MCP client to connect via SSE
Add an SSE-based MCP server entry pointing to the running container. Most modern MCP clients support SSE transport with a URL.
{
"mcpServers": {
"graphiti": {
"url": "http://localhost:8000/sse"
}
}
}Graphiti Examples
Client configuration
SSE-based MCP client configuration for Cursor or Claude Desktop connecting to the local Graphiti server.
{
"mcpServers": {
"graphiti": {
"url": "http://localhost:8000/sse"
}
}
}Prompts to try
Example prompts showing how to store and retrieve knowledge graph facts through the AI assistant.
- "Remember that the authentication service uses JWT tokens with a 24-hour expiry and store this as a fact in the knowledge graph"
- "Search the knowledge graph for everything we know about the payment module architecture"
- "Add an entity for our new microservice called 'inventory-api' and link it to the 'orders-service' entity"
- "What decisions have we stored about the database schema for the users table?"Troubleshooting Graphiti
Neo4j container fails to start or connection is refused
Run `docker compose logs neo4j` to see Neo4j startup errors. Make sure port 7687 is not already in use on your machine. If you changed NEO4J_PASSWORD in .env, ensure the value matches in both the neo4j and graphiti service sections of docker-compose.yml.
Embedding or LLM calls fail with authentication errors
Verify that OPENAI_API_KEY is correctly set in your .env file and that the Docker Compose environment is picking it up (`docker compose config` will show resolved env values). The key must have access to both the chat and embeddings endpoints.
MCP client cannot connect to the SSE endpoint
Confirm the server is running with `docker compose ps` and that port 8000 is mapped and accessible. Some MCP clients require the full SSE URL to end with `/sse` — use `http://localhost:8000/sse` exactly as shown.
Frequently Asked Questions about Graphiti
What is Graphiti?
Graphiti is a Model Context Protocol (MCP) server that mcp server for graphiti It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Graphiti?
Install via npm with the command: npx -y graphiti. Then add the server configuration to your AI client's JSON config file (e.g., claude_desktop_config.json or .cursor/mcp.json).
Which AI clients work with Graphiti?
Graphiti works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Graphiti free to use?
Yes, Graphiti is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
Graphiti Alternatives — Similar Knowledge & Memory Servers
Looking for alternatives to Graphiti? Here are other popular knowledge & memory servers you can use with Claude, Cursor, and VS Code.
MemPalace
★ 52.6kA local AI memory system that stores all conversations verbatim and organizes them into navigable structures. It provides 19 MCP tools for AI assistants to search and retrieve past decisions, debugging sessions, and architecture debates automatically
Kratos
★ 25.7k🏛️ Memory System for AI Coding Tools - Never explain your codebase again. MCP server with perfect project isolation, 95.8% context accuracy, and the Four Pillars Framework.
Context Mode
★ 15.4kAn MCP server that preserves LLM context by intercepting large data outputs and returning only concise summaries or relevant sections. It enables efficient sandboxed code execution, file processing, and documentation indexing across multiple programm
Memu
★ 13.7kMemory for 24/7 proactive agents like OpenClaw.
MemOS
★ 9.3kMemOS (Memory Operating System) is a memory management operating system designed for AI applications. Its goal is: to enable your AI system to have long-term memory like a human, not only remembering what users have said but also actively invoking, u
Everos
★ 5.4kBuild, evaluate, and integrate long-term memory for self-evolving agents.
Browse More Knowledge & Memory MCP Servers
Explore all knowledge & memory servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Graphiti in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Graphiti?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.