MSSQL MCP Server

Example Use Cases

• 1

  Inspect database table schemas and metadata to formulate accurate SQL queries or provide context-aware answers.

• 2

  Execute read queries and retrieve result sets (first 100 rows) in CSV format for analysis or summarization.

• 3

  Run modification queries (INSERT/UPDATE/DELETE) and receive affected row counts while using a standardized MCP interface.

Description

This project implements a Model Context Protocol (MCP) server in Python to provide Language Models programmatic access to Microsoft SQL Server databases. It supports asynchronous operation, connection pooling via pyodbc, FastAPI endpoints, and Pydantic validation to safely list tables, read table data, and run SQL queries. Results are returned in CSV format (with limits) and the server includes logging, error handling, and environment-based configuration for secure deployment. It is distributed under the MIT license and intended as a self-hosted, developer-friendly integration for LLM agents.