NeoCoder

NeoCoder is an MCP server that connects AI assistants like Claude Desktop to Neo4j knowledge graphs and Qdrant vector databases for standardized development workflows and hybrid knowledge reasoning. Designed for software engineers, researchers, and technical teams, it transforms a graph database into a dynamic instruction manual and persistent project memory. By querying Neo4j, the system guides the AI through structured action templates, mandatory testing procedures, and change verification steps before recording task completion. Simultaneously, it pairs with Qdrant to perform semantic vector searches and contextual similarity matching, unifying structured relationships with unstructured context. The server includes specialized process management features, such as graceful signal handling, connection pooling, and zombie process cleanup to maintain stability during intensive operations. Through tools like get_cleanup_status and check_connection, developers can inspect system health while ensuring that LLMs strictly adhere to repository-defined guidelines rather than unverified assumptions.

Category: AI Memory & Context

Tags: coding, graph database, knowledge-graph, neo4j, workflow

Visit NeoCoder

How to install and configure NeoCoder

Follow these steps to install and register NeoCoder with Claude Desktop: 1. Ensure you have Python 3.10+, uv, Docker, and a running Neo4j instance. 2. Clone the repository and navigate to its directory: bash git clone https://github.com/angrysky56/NeoCoder-neo4j-ai-workflow.git cd NeoCoder-neo4j-ai-workflow 3. Set up the virtual environment and install dependencies: bash uv venv source .venv/bin/activate uv pip install -e '.[dev,docs,gpu]' 4. Start Qdrant with Docker: bash docker run -p 6333:6333 -p 6334:6334 -v "$(pwd)/qdrant_storage:/qdrant/storage:z" qdrant/qdrant 5. Open your claude-app-config.json file and add the server configuration under mcpServers: json { "mcpServers": { "neocoder": { "command": "uv", "args": [ "--directory", "/your-path-to/NeoCoder-neo4j-ai-workflow", "run", "mcp_neocoder" ], "env": { "NEO4J_URL": "bolt://localhost:7687", "NEO4J_USERNAME": "neo4j", "NEO4J_PASSWORD": "your-neo4j-password-here", "NEO4J_DATABASE": "neo4j", "LOG_LEVEL": "INFO", "MCP_TRANSPORT": "stdio", "PYTHONUNBUFFERED": "1" } } } } 6. Replace the directory path and database credentials with your actual values, then restart Claude Desktop.

What you can do with NeoCoder

• Enforcing standardized coding and testing workflows stored as action templates within a Neo4j graph. - Querying Qdrant vector storage alongside Neo4j to perform hybrid semantic and relational knowledge reasoning. - Recording completed coding tasks, verification outcomes, and workflow execution nodes directly to graph memory. - Managing server processes and memory using built-in diagnostic tools like get_cleanup_status and check_connection. - Ingesting documentation and academic papers into interconnected knowledge graphs complete with citation tracking.

Key facts

• https://github.com/angrysky56/NeoCoder-neo4j-ai-workflow
• AI Memory & Context, Databases & Data Stores, Developer Tools & Code Intelligence
• coding, graph database, knowledge-graph, neo4j, workflow

Part of MCP Servers

What is NeoCoder?

NeoCoder is a Model Context Protocol server that bridges AI assistants to Neo4j knowledge graphs and Qdrant vector databases. It provides structured action templates, semantic retrieval, and persistent project memory to guide automated coding workflows.

Which MCP clients work with NeoCoder?

NeoCoder is designed to work with Claude Desktop using the stdio transport layer. Any MCP-compliant client that supports executing local commands via uv or Python and allows configuring environment variables can connect to it.

What databases are required to run NeoCoder?

You need a running Neo4j instance to handle structured knowledge graphs, action templates, and execution nodes, along with a Qdrant vector database to support semantic context search and hybrid reasoning.

How do I monitor the health of the NeoCoder server?

The server exposes diagnostic tools such as check_connection to verify Neo4j database connectivity and permissions, as well as get_cleanup_status to monitor background tasks, memory management, and process resource usage.