FastAPI server using CrewAI and Neo4j MCP

This is a simple FastAPI server that uses CrewAI and Neo4j MCP to process commands for creating or reading data in a Neo4j graph database.

Requirements

• Poetry for virtual env and dependency management
• OpenAI Key for LLM inference
• Running Neo4j database

Setup

1. Copy the sample.env file to .env and fill in the values.
2. Run poetry install

Run the crew_manager as a script

poetry run python crew_manager.py

Start FastAPI Server

poetry run uvicorn main:app --reload --port 4000

Interactive docs will be accessible at: http://127.0.0.1:8000/docs

License

MIT License