home / mcp / sql analysis mcp server
Provides an MCP server that analyzes SQL files and supports sampling for MCP clients.
Configuration
View docs{
"mcpServers": {
"jjdelorme-sql-analysis-mcp-server": {
"command": "uv",
"args": [
"run",
"mcp_server.py",
"--",
"--sql_path",
"/path/to/your/sql/files"
]
}
}
}You can run and test a Model Context Protocol (MCP) server that provides model-context data to clients. This server demonstrates how to expose data through MCP and how clients can sample and interact with it during development and testing.
Start the server in development to connect with an MCP client or inspector. In normal operation, an MCP client will query the server for data and actions defined by the MCP contract. You can run it locally, inspect its behavior through the MCP Inspector tool, and observe how sampling and context retrieval work within your client applications.
Prerequisites you need to prepare before installing and running the server are listed here.
# Prerequisites
# 1) Install Python 3.x
# 2) Install uv (Python package installer resolver)
# 3) Install fastmcp (MCP server/client toolkit)
# 4) Install Node.js and npm (to use npx for tooling)
# Install dependencies
uv syncDevelopment can be inspected with the MCP Inspector, which runs a browser-based interface to interact with the running server. In production, you expose the server so an MCP client can connect and exchange data.
# Development run pattern shown for inspector-based development
npx @modelcontextprotocol/inspector uv run mcp_server.py -- --sql_path /path/to/your/sql/filesLaunches the MCP Inspector web UI to explore and interact with the running MCP server.
Executes the MCP server using the uv command and the mcp_server.py script, enabling MCP communications.
Generates a client JSON configuration via fastmcp for connecting MCP clients to this server.