MCP Neo4j Knowledge Graph

🚀 MCP Knowledge Graph Project Documentation

The MCP Knowledge Graph project aims to build and manage a knowledge graph in the medical field using the Neo4j graph database. It provides a variety of functional modules for data management and analysis.

🚀 Quick Start

The MCP (Medical Cybernetics Platform) Knowledge Graph project is designed to construct and manage a knowledge graph in the medical domain with the Neo4j graph database. It offers a rich set of functional modules, including data import, entity operations, relationship management, and schema insights.

✨ Features

Core Components

1. MCP Client: Interacts with the Neo4j database.
2. Configuration Tool: Loads and validates configuration files.
3. Neo4j Connection Test: Ensures database connectivity and basic query functionality.
4. Integration Test: Verifies the integrity and compatibility of the entire system.

Functional Modules

1. MCP Client Features

• Start the Server: Run the command task run to start the MCP service.
• Tool List: Access the /tools interface to get information about available tools.
• Schema Insights: Call the /schema interface to obtain details of the database schema, including node labels, relationship types, and their attributes.

2. MCP Configuration Management

• Configuration File Loading: Validate the correctness of the config.yaml file.
• Connection Test: Use the official MCP SDK to test the connection status with the database.
• Tool Availability Check: Ensure that all required tools are installed.

3. Neo4j Database Connection

• Connectivity Verification: Test whether the database is reachable through basic queries.
• Environment Configuration: Check and confirm that the operating environment settings meet the requirements.

💻 Usage Examples

Basic Usage

Test Scripts

1. mcp_neo4j_knowledge_graph/test_mcp_client.py:

   • Verify the server startup status.
   • Test the tool list interface.
   • Validate the schema insights function.
   • Test entity creation operations.

   task test-client  # Only run client tests
   

2. mcp_neo4j_knowledge_graph/test_mcp_config.py:

   • Test the configuration file loading logic.
   • Verify the MCP SDK connection status.
   • Ensure all tools are available.

   task test-config  # Only run configuration tests
   

3. mcp_neo4j_knowledge_graph/test_neo4j_connection.py:

   • Test database connectivity.
   • Verify basic query functionality.
   • Confirm that the environment configuration is correct.

   task test-db  # Only run database tests
   

Task Commands

# Start the service
task run

# Execute tests
task test

# Format the code
task format

# Run checks
task lint

# Start the development environment
task dev

Advanced Usage

Using Poetry

# Format the code
poetry run black .

# Sort imports
poetry run isort .

# Run code checks
poetry run flake8

# Execute tests
poetry run pytest

📚 Documentation

Error Handling

The server implements a comprehensive error handling mechanism covering the following scenarios:

• Database connection issues
• Invalid queries
• Missing nodes
• Incorrect request formats
• Schema validation failures
• Relationship creation exceptions
• Entity update conflicts

All errors return error messages compliant with the MCP protocol.

Docker Configuration

The Neo4j container is configured as follows:

• Ports: 7474 (HTTP) and 7687 (Bolt)
• Default Credentials: neo4j/password
• APOC Plugin Enabled
• File Import/Export Allowed
• Health Check Configured

These settings can be modified in the docker-compose.yml file.

Task Command Reference

# Start the service
task run

# Execute tests
task test

# Format the code
task format

# Run checks
task lint

# Start the development environment
task dev