MCP Memgraph

Memgraph MCP Server is a lightweight implementation of the Model Context Protocol (MCP) server, aiming to connect the Memgraph graph database with large language models (LLMs) and provide Cypher query execution and schema information acquisition functions.

Database Artificial intelligence chatbots #Graph database #LLM connection #Lightweight service .Python

rating : 2.5 points

downloads : 6.6K

update time : 2025-04-28

Open Site

What is Memgraph MCP Server?

Memgraph MCP Server is a lightweight service based on the Model Context Protocol. It acts as a bridge between the Memgraph graph database and AI assistants (such as Claude). Through a standardized protocol, it allows users to interact with the graph database using natural language.

How to use Memgraph MCP Server?

Just complete three simple steps: 1) Start the MCP service. 2) Configure the AI client. 3) Start natural language queries. The system will automatically convert your requests into professional database queries.

Applicable scenarios

It is particularly suitable for non - technical users such as data analysts and product managers to quickly query graph data. It is also suitable for developers to integrate graph database capabilities into AI applications.

Main features

Cypher query execution

Execute Cypher query statements directly through the run_query() tool

Intelligent schema acquisition

Automatically obtain database structure information using get_schema() to help LLMs understand the data model

Advantages

No need to learn complex query languages. You can operate the database using natural language

Lightweight design for quick deployment and integration

Supports standard protocols of mainstream AI platforms (such as Claude)

Limitations

Currently only supports the Memgraph database

Requires pre - configuration of the schema - info - enabled parameter

Support for the Windows platform is still being improved

How to use

Prepare the Python environment

Use the uv tool to create a virtual environment and install dependencies

Start the MCP service

Run the service in the project directory

Configure the AI client

Modify the Claude configuration file to add MCP server information

Start Memgraph

Run a Memgraph instance with the MAGE extension using Docker

Usage examples

Data exploration

Quickly explore data relationships without knowing the database structure

Business analysis

Perform complex business analysis queries without writing code

Frequently Asked Questions

Why do I need to set schema - info - enabled=True?

How can I confirm that the service has been started correctly?

Which AI clients are supported?

Related resources

Quick start video

A 5 - minute quick configuration demonstration

Memgraph Lab dataset

An example dataset for testing

AI Toolkit project

Memgraph's official AI tool collection

🚀 Memgraph MCP Server

The Memgraph MCP Server is a lightweight implementation of the Model Context Protocol (MCP), designed to connect Memgraph to large language models (LLMs).

🚀 Quick Start

📹 Quick Start Video for Memgraph MCP Server

1. Run the Memgraph MCP Server

Install uv and create a venv environment: uv venv. Activate the virtual environment using .venv\Scripts\activate.
Install dependencies: uv add "mcp[cli]" httpx
Run the Memgraph MCP Server: uv run server.py.

2. Run the MCP Client

Install Claude Desktop.
Add the Memgraph server to the Claude configuration:

MacOS/Linux

code ~/Library/Application\ Support/Claude/claude_desktop_config.json

Windows

code $env:AppData\Claude\claude_desktop_config.json

Configuration example:

{
    "mcpServers": {
      "mpc-memgraph": {
        "command": "/Users/katelatte/.local/bin/uv",
        "args": [
            "--directory",
            "/Users/katelatte/projects/mcp-memgraph",
            "run",
            "server.py"
        ]
     }
   }
}

⚠️ Important Note

You may need to provide the full path to the uv executable (in the command field). You can obtain it by running which uv (MacOS/Linux) or where uv (Windows). Make sure to pass the absolute path of the Memgraph server.

3. Talk to the Database

Run Memgraph MAGE:
```
docker run -p 7687:7687 memgraph/memgraph-mage --schema-info-enabled=True
```
Set the configuration parameter --schema-info-enabled to True to allow the LLM to execute the SHOW SCHEMA INFO query.
Open Claude Desktop and view the Memgraph tools and resources. Try using them! (You can load sample data from the Memgraph Lab dataset)

🔧 Tools

run_query()

Run Cypher queries against Memgraph.

🗃️ Resources

get_schema()

Get the schema information of Memgraph (Prerequisite: --schema-info-enabled=True).

🗺️ Roadmap

The Memgraph MCP Server is still in its early stages. We are actively developing its features to make it easier to integrate Memgraph into modern AI workflows. In the near future, we will release a TypeScript version of the server to better support JavaScript-based environments. Additionally, we plan to migrate this project to our core AI Toolkit repository, where it will be part of a unified open-source toolkit along with other tools and integrations such as LangChain, LlamaIndex, and MCP, making it more seamless to build graph-driven applications and intelligent agents centered around Memgraph.