qdrant-neo4j-crawl4ai

MCP server combining Qdrant vector search, Neo4j knowledge graphs, and Crawl4AI web intelligence with agentic RAG capabilities. FastMCP 2.0 architecture with enterprise security, monitoring, and Kubernetes deployment. AI/ML engineering powerhouse.

Installation

Run this command in your terminal to add the MCP server to Claude Code.

Run in terminal:

View docs

Command

claude mcp add --transport stdio bjornmelin-qdrant-neo4j-crawl4ai-mcp docker run -i qdrant-neo4j-crawl4ai-mcp \
  --env NEO4J_URI="bolt://localhost:7687" \
  --env NEO4J_USER="neo4j" \
  --env QDRANT_URL="http://localhost:6333" \
  --env CORS_ORIGINS="https://your-domain.com" \
  --env JWT_SECRET_KEY="your-secure-secret-key" \
  --env NEO4J_PASSWORD="password" \
  --env MCP_SERVER_HOST="localhost" \
  --env MCP_SERVER_PORT="8000" \
  --env RATE_LIMIT_PER_MINUTE="100"

How to use

This MCP server provides an agentic retrieval-augmented generation platform that combines a Qdrant vector store, a Neo4j graph memory system, and Crawl4AI web intelligence. It exposes a single Model Context Protocol (MCP) interface that lets clients perform vector searches, graph memory queries, and web-crawled content extraction through a unified API. Use the REST endpoints documented in the API docs to perform tasks such as semantic search against your embedded data, retrieve memories or relationships from the graph, and fetch real-time information from the web. The architecture supports autonomous orchestration, routing queries to the most relevant service and fusing results for a coherent answer.

How to install

Prerequisites:

Docker and Docker Compose (for containerized deployment)
Git
Optional: Python 3.11+ and uv if you prefer local development

Installation steps:

Clone the repository: git clone https://github.com/BjornMelin/qdrant-neo4j-crawl4ai-mcp.git cd qdrant-neo4j-crawl4ai-mcp
Using Docker (recommended for production):
- Build the Docker image: docker build -t qdrant-neo4j-crawl4ai-mcp .
- Run with Docker (example): docker run -p 8000:8000 qdrant-neo4j-crawl4ai-mcp
- Ensure required environment variables are set (see the env section below).
Local development (uv/pipx alternative):
- Install dependencies and run the MCP locally: uv sync cp .env.example .env # then edit with your config uv run python -m qdrant_neo4j_crawl4ai_mcp
Docker Compose (optional):
- If a docker-compose file is provided, start services: docker-compose up -d

Prerequisites to configure before first run:

Set MCP_SERVER_HOST, MCP_SERVER_PORT, and JWT_SECRET_KEY
Configure QDRANT_URL, NEO4J_URI, NEO4J_USER, NEO4J_PASSWORD
Adjust RATE_LIMIT_PER_MINUTE and CORS_ORIGINS as needed

Additional notes

Tips and common issues:

Ensure Qdrant and Neo4j databases are reachable at the configured URLs before starting the MCP server.
If you encounter authentication issues, verify JWT_SECRET_KEY and token handling in your client.
For local development, use uv run to start the MCP and ensure Python 3.11+ is installed.
When using Docker, keep container networking in mind; map ports appropriately and expose the required endpoints (default 8000).
The API docs (Swagger UI and ReDoc) are available once the server is running at http://localhost:8000/docs and http://localhost:8000/redoc.
Monitor health checks and logs to diagnose issues quickly; enable monitoring and observability features as described in the docs.

Related MCP Servers

kubectl

839

A Model Context Protocol (MCP) server for Kubernetes. Install: npx kubectl-mcp-server or pip install kubectl-mcp-server

mcpcan

715

MCPCAN is a centralized management platform for MCP services. It deploys each MCP service using a container deployment method. The platform supports container monitoring and MCP service token verification, solving security risks and enabling rapid deployment of MCP services. It uses SSE, STDIO, and STREAMABLEHTTP access protocols to deploy MCP。

mcp-for-argocd

335

An implementation of Model Context Protocol (MCP) server for Argo CD.

k8s

148

Manage Your Kubernetes Cluster with k8s mcp-server

MCP-PostgreSQL-Ops

137

🔍Professional MCP server for PostgreSQL operations & monitoring: 30+ extension-independent tools for performance analysis, table bloat detection, autovacuum monitoring, schema introspection, and database management. Supports PostgreSQL 12-17.

magg

130

Magg: The MCP Aggregator