pyodbc MCP Server

A Model Context Protocol (MCP) server that provides read-only access to Microsoft SQL Server databases using Windows Authentication.

Built for environments where:

• 🔐 Windows Authentication is required (no username/password storage)
• 🛡️ Read-only access is mandated by IT security policy
• 🖥️ SQL Server is accessed from Windows workstations
• 🤖 AI assistants need safe database access (Claude Code, etc.)

🆕 What's New in v0.4.0 (Production Ready!)

Version 0.4.0 marks the first production-ready release with:

• ✅ Published on PyPI - Simple pip install pyodbc-mcp-server
• ✅ Comprehensive Testing - 193 tests, 83% code coverage
• ✅ Professional Configuration - CLI args, TOML files, environment variables
• ✅ Robust Error Handling - Typed exceptions with automatic retry logic
• ✅ Complete Documentation - 4,893 lines across 5 guides (API, Config, Troubleshooting, Examples, Development)

Upgrade from pre-release versions: pip install --upgrade pyodbc-mcp-server

Features

• Windows Authentication - Uses Trusted_Connection via pyodbc, no credentials to manage
• Read-only by design - Only SELECT queries allowed, dangerous keywords blocked
• Comprehensive error handling - Typed exceptions with retry logic for transient failures
• Row limiting - Prevents accidental large result sets (configurable, max 10,000)
• Schema exploration - 10 tools for tables, views, indexes, constraints, relationships, and more
• MCP resources - 5 URI-based endpoints for quick data access
• Configurable - CLI arguments, TOML config files, or environment variables
• Production-ready - Query timeouts, connection retries, comprehensive logging

Available Tools

See API Reference for complete documentation.

Installation

🚀 Quick Install (Recommended)

The package is available on PyPI and ready for production use:

pip install pyodbc-mcp-server

That's it! The server is now available as the pyodbc-mcp-server command.

📦 Latest Release: v0.4.0 (Production Ready) 📚 Full Documentation: docs/ 🔗 PyPI Package: pypi.org/project/pyodbc-mcp-server

Prerequisites

• Python 3.10+ (tested on 3.10, 3.11, 3.12)
• Windows with ODBC Driver 17+ for SQL Server
• Network access to your SQL Server
• Windows domain account with SELECT permissions on target database

Install ODBC Driver (if needed)

If you don't have the ODBC driver installed, download Microsoft ODBC Driver 17 for SQL Server.

Alternative: Install from Source

For development or the latest unreleased features:

git clone https://github.com/jjones-wps/pyodbc-mcp-server.git
cd pyodbc-mcp-server
pip install -e ".[dev]"

Configuration

Quick Configuration

The server supports three configuration methods (in priority order):

1. CLI Arguments (highest priority)
2. TOML Configuration File
3. Environment Variables (lowest priority)

Minimal setup with defaults:

mssql-mcp-server

Using TOML config file:

# Create config from example
cp config.example.toml config.toml

# Edit config.toml, then run
mssql-mcp-server --config config.toml

Override specific settings:

mssql-mcp-server --config config.toml --database AdventureWorks --query-timeout 120

See Configuration Guide for complete documentation.

Configuration Parameters

Environment Variables

Claude Code Configuration

Quick Install via CLI (Recommended)

Since the package is on PyPI, adding to Claude Code is simple:

claude mcp add mssql --transport stdio -e MSSQL_SERVER=your-server -e MSSQL_DATABASE=your-database -- pyodbc-mcp-server

Note: Claude Code will automatically use the PyPI package. No need to specify installation paths!

With all environment variables:

claude mcp add mssql --transport stdio \
  -e MSSQL_SERVER=your-sql-server \
  -e MSSQL_DATABASE=your-database \
  -e ODBC_DRIVER="ODBC Driver 17 for SQL Server" \
  -- pyodbc-mcp-server

Scope options:

# User scope - available across all your projects (default)
claude mcp add mssql --transport stdio -e MSSQL_SERVER=your-server -e MSSQL_DATABASE=your-db -- pyodbc-mcp-server

# Project scope - shared with team via .mcp.json (checked into version control)
claude mcp add mssql --transport stdio --scope project -e MSSQL_SERVER=your-server -e MSSQL_DATABASE=your-db -- pyodbc-mcp-server

Verify installation:

claude mcp list          # List all configured servers
claude mcp get mssql     # Show details for this server

Manual Configuration (Alternative)

Add to your ~/.claude.json (or %USERPROFILE%\.claude.json on Windows):

{
  "mcpServers": {
    "mssql": {
      "type": "stdio",
      "command": "pyodbc-mcp-server",
      "env": {
        "MSSQL_SERVER": "your-sql-server",
        "MSSQL_DATABASE": "your-database"
      }
    }
  }
}

Note for Windows users: If you encounter issues, try wrapping with cmd /c:

"command": "cmd",
"args": ["/c", "pyodbc-mcp-server"],

Claude Desktop Configuration

Add to your Claude Desktop config (%APPDATA%\Claude\claude_desktop_config.json):

{
  "mcpServers": {
    "mssql": {
      "command": "pyodbc-mcp-server",
      "env": {
        "MSSQL_SERVER": "your-sql-server",
        "MSSQL_DATABASE": "your-database"
      }
    }
  }
}

Note: Since v0.4.0, pyodbc-mcp-server is installed as a command-line tool. For older installations, use "command": "python", "args": ["-m", "mssql_mcp_server"].

Usage Examples

Once configured, you can ask Claude to:

Explore Schema

"List all tables in the dbo schema"
"Describe the structure of the customers table"
"What are the foreign key relationships for the orders table?"

Query Data

"Show me the first 10 rows from the products table"
"Find all orders from the last 30 days"
"What are the top 5 customers by total order value?"

Analyze Relationships

"Find all tables that reference the customer table"
"Show me the relationship between orders and order_lines"

Security

This server is designed with security as a primary concern:

Read-Only Enforcement

• Only queries starting with SELECT are allowed
• Dangerous keywords are blocked even in subqueries:
  • INSERT, UPDATE, DELETE, DROP, CREATE, ALTER
  • EXEC, EXECUTE, TRUNCATE, GRANT, REVOKE, DENY
  • BACKUP, RESTORE, SHUTDOWN, DBCC

Windows Authentication

• Uses Trusted_Connection=yes - no passwords stored or transmitted
• Leverages existing Windows domain security
• Your database permissions are enforced by SQL Server

Row Limiting

• Default limit: 100 rows per query
• Maximum limit: 10,000 rows per query
• Prevents accidental retrieval of large datasets

Error Handling & Retry Logic

• Typed Exceptions: ConnectionError, QueryError, SecurityError, ValidationError, TimeoutError
• Automatic Retries: Transient errors (connection failures, timeouts, deadlocks) are retried with exponential backoff
• Configurable Timeouts: Separate timeouts for connection and query execution
• Consistent Error Format: All errors returned as JSON with error code, message, and details

Documentation

Comprehensive documentation is available in the docs/ directory:

Development

Quick Start

# Clone repository
git clone https://github.com/jjones-wps/pyodbc-mcp-server.git
cd pyodbc-mcp-server

# Install with development dependencies
pip install -e ".[dev]"

# Run tests
pytest

# Check coverage
pytest --cov=src/mssql_mcp_server --cov-report=html

# Format code
black src tests && isort src tests

# Type check
mypy src

# Lint
ruff check src tests

Running Locally

# Using environment variables
export MSSQL_SERVER=your-server
export MSSQL_DATABASE=your-database
python -m mssql_mcp_server

# Using config file
cp config.example.toml config.toml
# Edit config.toml
mssql-mcp-server --config config.toml

# Validate configuration
mssql-mcp-server --config config.toml --validate-only

See Development Guide for architecture, testing patterns, and contribution workflow.

Contributing

Contributions are welcome! Please see the Development Guide for:

• Development setup
• Architecture overview
• Testing guide
• Code style requirements
• Adding new tools
• Release process

Quick contribution checklist:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/my-feature)
3. Make changes and add tests
4. Ensure tests pass and coverage doesn't decrease
5. Format code (black src tests && isort src tests)
6. Submit a pull request

License

MIT License - see LICENSE file.

Acknowledgments

• Built with FastMCP for MCP protocol handling
• Uses pyodbc for SQL Server connectivity
• Inspired by the need for safe AI access to enterprise databases

Related Projects

• MCP Specification
• Claude Code
• FastMCP