MSSQL MCP Server

What is MSSQL MCP Server?

An MCP server for read-only Microsoft SQL Server interactions, providing safe, read-only access to MSSQL databases through the Model Context Protocol. It is designed for developers who need to explore database schemas and run validated SELECT queries without modifying data.

How to use MSSQL MCP Server?

Install the package with pip install -e ., create a .env file with connection details (e.g., MSSQL_SERVER, MSSQL_DATABASE, MSSQL_USERNAME, MSSQL_PASSWORD), then run the server in development mode with mcp dev src/mssql_mcp/server.py, integrate with Claude Desktop via mcp install src/mssql_mcp/server.py --name "MSSQL Server", or execute directly with python -m mssql_mcp.server.

Key features of MSSQL MCP Server

• Read-only operations: only SELECT, DECLARE, and SET allowed
• Schema exploration: browse databases, tables, columns, and indexes
• SQL injection protection and query validation
• Secure connection handling with proper cleanup
• Accessible via MCP resources and tools

Use cases of MSSQL MCP Server

• Developers exploring MSSQL schemas safely without write access
• Building MCP-enabled applications requiring read-only database access
• Safe integration of MSSQL into AI assistants for schema browsing
• Database teams providing read-only access for analysis or reporting

FAQ from MSSQL MCP Server

What operations are allowed?

Only SELECT, DECLARE, and SET (for local variables) statements; no DDL or DML operations are permitted.

What are the system requirements?

Python 3.9+, access to a Microsoft SQL Server instance, and the pymssql library (installed automatically with the package).

How is the database connection configured?

Connection details are set in a .env file in the current directory, using environment variables like MSSQL_SERVER, MSSQL_DATABASE, MSSQL_USERNAME, and MSSQL_PASSWORD.

Does the server support authentication via credentials?

Yes, it uses MSSQL_USERNAME and MSSQL_PASSWORD environment variables, with secure connection handling.

What resources are available for schema exploration?

Resources such as db://schemas, db://tables/{database}, db://schema_tables/{schema_name}, db://columns/{schema_name}/{table_name}, and db://indexes/{database}/{schema_name}/{table_name} are provided.