Knowledge MCP

🚀 Knowledge - MCP Server

This is an MCP server that provides knowledge from a Neo4J database. It solves the problem of retrieving knowledge from Neo4J and can be used as a knowledge source for relevant clients.

🚀 Quick Start

Run Neo4J in a Docker Container

Run the Neo4J Docker container:

docker run \
    --restart always \
    --publish=7474:7474 \
    --publish=7687:7687 \
    neo4j:latest

Log in to Neo4J at http://localhost:7474/ in your browser and set a new password for the user neo4j.

Run the Data Loader

Run the data loader project to save some documents to Neo4J. These documents will be used by the MCP server to answer questions.

💡 Usage Tip

Remember to modify the source code to point to the Neo4J instance you are running and set the correct Neo4J username and password.

Build the MCP Server

Build the server to generate the required binary files.

Configure the MCP Server in Cursor

Go to Cursor settings and add a new MCP server.

Remember to set YOUR_PATH, YOUR_OPENAI_API_KEY, and YOUR_NEO4J_PASSWORD to the correct values.

{
  "mcpServers": {
    "KnowledgeBase": {
      "type": "stdio",
      "command": "YOUR_PATH/knowledge-mcp/src/KnowledgeMcp.Server/bin/Debug/net9.0/KnowledgeMcp.Server",
      "args": [
        "YOUR_OPEN_AI_API_KEY",
        "neo4j://localhost:7687",
        "neo4j",
        "YOUR_NEO4J_PASSWORD"
      ]
    }
  }
}

⚠️ Important Note

The guides provided in this document are of an example nature, not real, and are only for the purpose of demonstrating how to extract knowledge from Neo4J. This is a prototype and is not ready for production use!

💡 Usage Tip

You can use any MCP - supported client, such as Claude Desktop. One possible improvement is to utilize Neo4J's response format to obtain document references for answering questions.