Steampipe

🚀 Steampipe MCP Server

This is a simple Steampipe MCP server. It acts as a bridge between your AI model and the Steampipe tool.

🚀 Quick Start

✨ Features

This server serves as an intermediary, enabling seamless interaction between AI models and the Steampipe tool.

📦 Installation

Prerequisites

• Install Python 3.10 or a higher version.
• Install uv (my favorite) and mcp[cli].
• Install and configure Steampipe to ensure it runs properly.
• Configure Steampipe plugins (e.g., the github plugin) and prepare relevant credentials (e.g., place tokens in ~/.steampipe/config/github.spc).
• Use any LLM that supports MCP. Here, Claude is used.
• Install Node.js and npx (for running the MCP inspector and possibly certain MCP servers).

💻 Usage Examples

Running the MCP Inspector

This is an excellent tool for testing whether your MCP server is working correctly.

• Run the inspector: npx -y @modelcontextprotocol/inspector uv --directory . run steampipe_mcp_server.py
• A browser window will open the MCP inspector UI (usually at http://localhost:XXXX).
• Wait for the left - hand panel to display the "Connected" status.
• Go to the Tools tab.
• You should see the run_steampipe_query tool and its description in the tool list.
• Click on the tool name.
• In the "Parameters" JSON input field, enter a valid Steampipe query:

{
  "query": "select name, fork_count from github_my_repository"
}

• Execute and view the JSON results.

Running the Tool

It's very straightforward. Just run the inspector and ensure the tool is working properly. Then add the server configuration to the corresponding LLM and select the tool from the LLM.

📚 Documentation

Troubleshooting

• If the tool is not found in the inspector, there is an issue with the @mcp.tool() decorator.
• Execution errors - Check the "Results" in the inspector and the server logs (stderr). Is Steampipe running? Are there SQL errors? Timeouts? JSON parsing errors? Adjust the Python script.

⚠️ Important Note

Claude blindly executes your SQL queries, so there may be the possibility of arbitrary SQL query generation and execution in this POC.