Neo4j MCP Chainlit

🚀 Neo4j MCP for Chainlit

A proof-of-concept project demonstrating the integration of the Neo4j MCP server with Chainlit (MCP host) and Claude LLM (Anthropic API).

🚀 Quick Start

This project creates an interactive chat interface for querying a Neo4j database using natural language. It combines:

• Chainlit to provide a web interface
• Neo4j's MCP (Model Context Protocol) for database access
• Anthropic's Claude as an LLM for natural language understanding

Prerequisites

1. Clone the repository:

git clone https://github.com/Abhid14/neo4j-mcp-chainlit.git
cd neo4j-mcp-chainlit

2. Install uv on your system:

curl -LsSf https://astral.sh/uv/install.sh | sh

For other installation options, refer to the uv documentation.

Setup

1. Create a Python virtual environment:

uv venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

2. Install the dependencies:

uv pip install -r requirements.txt

3. Configure the environment variables:

cp .env.example .env

4. Add your Anthropic API key to the .env file:

ANTHROPIC_API_KEY=your_api_key_here

Run the Application

Start the Chainlit application:

chainlit run app.py -w

Configure the MCP Connection

1. In the Chainlit application interface, configure the MCP connection.
2. Use the example Neo4j database:
   • Set the MCP connection to stdio mode.
   • Name it neo4j-mcp-demo.
   • Set the command to:

     /path/to/uv/binary/uvx mcp-neo4j-cypher --db-url neo4j+s://demo.neo4jlabs.com --user recommendations --password recommendations
     

💻 Usage Examples

The application uses the Neo4j example database (movie graph) to demonstrate the natural language query functionality.

Try asking the following questions:

• "Which movies did Tom Hanks act in?"
• "Show the relationships between actors and directors."
• "Find all movies released after 2010."