machbase

The Machbase MCP Server enables natural language based database operations by connecting Claude Desktop to Machbase Neo. When you describe what you want in Claude Desktop, the MCP server translates your request into Machbase SQL or Machbase Neo commands and executes them, returning results and explanations. Core capabilities include real-time database integration such as table creation, data manipulation, and executing complex queries, as well as time-series analysis for IoT and log data. It also leverages Claude Desktop’s built-in features, including documentation search for Machbase and AI-assisted explanations, to help you craft and understand database operations without manually writing SQL. Typical workflows involve asking questions about your Machbase schema, creating or updating tables, inserting or querying data, and retrieving results or summaries directly in Claude’s interface.

To use it, ensure Claude Desktop is configured to connect to this MCP server. Your natural language requests will be converted by the MCP server into Machbase commands and executed against Machbase Neo. Results and any generated SQL or explanations will be returned to Claude Desktop for presentation and further refinement if needed.

Prerequisites:

• A machine with Python installed (the README uses a Python-based setup).
• Access to Machbase Neo (and corresponding connection details: host, port, user, password).
• Claude Desktop installed and configured to connect via MCP.

Step-by-step installation:

1. Clone the repository containing the Machbase MCP server: git clone https://github.com/your-org/Machbase_MCP_Server.git cd Machbase_MCP_Server

2. Set up a Python environment (recommended):

   • Using Conda: conda create -n mcp python=3.11 conda activate mcp
   • Or using venv: python -m venv mcp-env source mcp-env/bin/activate # Linux/macOS mcp-env\Scripts\activate # Windows

3. Install dependencies: pip install -r requirements.txt

4. Claude Desktop and MCP Server configuration:

   • Create Claude Desktop config file at the following location (example): C:/Users/YourUser/AppData/Roaming/Claude/claude_desktop_config.json
   • Place Machbase.py or ensure the path to the script is correct in the mcp_config section.

5. Configure the MCP server in the Claude config (example): { "mcpServers": { "machbase": { "command": "C:/Path/Python/python.exe", "args": ["C:/Path/Machbase.py"], "env": { "MACHBASE_HOST": "localhost", "MACHBASE_PORT": "5654", "MACHBASE_USER": "sys", "MACHBASE_PASSWORD": "manager" } } } }

6. Run and test:

   • Start the MCP server via the configured command/args
   • Restart Claude Desktop
   • Test with natural language prompts like that shown in the examples to ensure correct translation and execution.

7. Troubleshooting:

   • Verify Machbase Neo is reachable at the specified host/port.
   • Ensure environment variables for Machbase credentials are correct.
   • Check Python dependencies if the script fails to import modules.

Tips and known notes:

• Ensure Machbase Neo is running and accessible from the machine running the MCP server.
• If Claude Desktop cannot reach Machbase, double-check the host/port values and network connectivity.
• The MCP server translates natural language into Machbase SQL/Neo commands; you can refine queries by iterating on your prompts.
• Store sensitive credentials securely; consider using environment variable management or secret vaults in production setups.
• If you modify Machbase.py, restart the MCP server and Claude Desktop to apply changes.
• If you see slow responses, check the complexity of the generated queries and Machbase performance metrics.