mcp-steampipe

mcp-steampipe

1
cloud_platforms

The Steampipe MCP is a server that bridges AI models with the Steampipe tool, requiring Python and Node.js installations along with proper configurations. It includes tools for testing connections and executing queries securely, although users must beware of potential security risks involving SQL queries.

Steampipe MCP

This is a simple Steampipe MCP server acting as a bridge between an AI model and the Steampipe tool.

Pre-requisites

  • Python 3.10+ installed.
  • Required packages: uv and mcp[cli].
  • Steampipe installed and configured with necessary credentials.
  • LLM supporting MCP, e.g., Claude.
  • Node.js and npx installed.

Running MCP Interceptor

Test if your MCP server is working using MCP Interceptor. It opens an MCP Inspector UI in the browser to view and execute queries. Ensure correct Steampipe query input.

Running the tool

To run the tool, ensure the interceptor is set and the tool is working. Add server configurations to LLM and select the tool accordingly.

TroubleShooting

  • Check if the @mcp.tool() decorator works.
  • For execution issues, inspect the MCP Inspector and server logs.

Security Risk: Be aware of security risks as the tool can execute arbitrary SQL queries using configured credentials.