JDBCX MCP Server

pydbcx-mcp is a Python implementation of MCP server for enabling communication with diverse data sources via JDBCX server.

Installation

Start JDBCX server

Starts the JDBCX server container. Check out here for more information.

# Start the server
docker run --rm --name bridge -d -p8080:8080 jdbcx/jdbcx server
# Test if the server if ready
curl -v 'http://localhost:8080/config'
# Check server logs
docker logs --tail=100 -f bridge
# Shutdown the server
docker stop bridge

Configure MCP server

To install JDBCX MCP server for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install @jdbcx/pydbcx-mcp --client claude

Alternatively, add the MCP server into your JSON config file.

Development/Unpublished Server Configuration

{
  "mcpServers": {
    "jdbcx": {
      "command": "uv",
      "args": [
        "--directory",
        "</path/to/your/pydbcx-mcp/dir>",
        "run",
        "pydbcx-mcp"
      ],
      "env": {
        "DEFAULT_QUERY_TIMEOUT_SECONDS": "30",
        "JDBCX_SERVER_URL": "http://localhost:8080/",
        "JDBCX_SERVER_TOKEN": "",
        "MAX_ROWS_LIMIT": "5000"
      }
    }
  }
}

Published Server Configuration

{
  "mcpServers": {
    "jdbcx": {
      "command": "uvx",
      "args": ["pydbcx-mcp"],
      "env": {
        "DEFAULT_DATA_FORMAT": "md"
      }
    }
  }
}

Published SSE Server Configuration

{
  "mcpServers": {
    "jdbcx": {
      "url": "http://localhost:8080/sse"
    }
  }
}

Note: remember to start the SSE server first by JDBCX_SERVER_URL=http://localhost:8080/ DEFAULT_QUERY_TIMEOUT_SECONDS=30 uvx pydbcx-mcp --transport sse.

Configuration

Configure the server using environment variables:

Note: It is highly recommended to enable access token in JDBCX server and configure JDBCX_SERVER_TOKEN accordingly for security reason.