MSSQL MCP Server

A Model Context Protocol (MCP) server for read-only Microsoft SQL Server interactions. This server provides safe, read-only access to MSSQL databases through the MCP protocol.

Features

• Read-only operations: Only SELECT, DECLARE, and SET (for local variables) queries are allowed
• Schema exploration: Browse databases, tables, columns, and indexes
• Safe querying: SQL injection protection and query validation
• Connection management: Secure connection handling with proper cleanup

Installation

1. Clone this repository:

git clone <repository-url>
cd mssql-mcp

2. Install dependencies:

pip install -e .

3. Set up environment variables:

cp .env.example .env
# Edit .env with your database connection details

Important: The server automatically loads environment variables from a .env file in the current directory. Make sure to create this file with your database connection details.

Configuration

Create a .env file with your MSSQL connection details:

MSSQL_SERVER=localhost
MSSQL_DATABASE=your_database
MSSQL_USERNAME=your_username
MSSQL_PASSWORD=your_password
MSSQL_PORT=1433
MSSQL_TIMEOUT=30
MSSQL_CHARSET=UTF-8
MSSQL_AUTOCOMMIT=true

Usage

Development Mode

mcp dev src/mssql_mcp/server.py

Claude Desktop Integration

mcp install src/mssql_mcp/server.py --name "MSSQL Server"

Direct Execution

python -m mssql_mcp.server

Available Resources

• db://schemas - List all schemas in the current database
• db://tables/{database} - List all tables in the current database
• db://schema_tables/{schema_name} - List tables in a specific schema
• db://columns/{schema_name}/{table_name} - Get column information for a table
• db://indexes/{database}/{schema_name}/{table_name} - Get index information for a table

Available Tools

• list_databases - List all accessible databases
• list_tables - List tables in the current database

Security

This server implements several security measures:

• Only SELECT, DECLARE, and SET (for local variables) statements are allowed
• SQL parsing and validation
• Parameterized queries where possible
• Connection string security
• No DDL or DML operations permitted

Requirements

• Python 3.9+
• Access to a Microsoft SQL Server instance
• pymssql library (automatically installed with dependencies)

License

MIT License