SQL Server

A Model Context Protocol (MCP) server that provides AI assistants with SQL Server database access capabilities.

- Execute SQL queries with automatic safety limits
- Database schema inspection and exploration
- Table statistics and performance monitoring
- Advanced search capabilities for tables and columns
- Table backup functionality
- Data insertion with conflict handling
- Query execution plan analysis
- Connection health monitoring

# Navigate to project directory cd C:\Users\benha\Desktop\sql-server-mcp # Create virtual environment python -m venv venv venv\Scripts\activate # Install dependencies pip install -r requirements.txt

Edit the.envfile with your SQL Server credentials:

SQL_SERVER_HOST=192.168.1.117 SQL_SERVER_DATABASE=EM_Data SQL_SERVER_USERNAME=benhg SQL_SERVER_PASSWORD=your_actual_password SQL_SERVER_PORT=1433

Add to your Claude Desktop configuration file:

- Windows:%APPDATA%\Claude\claude_desktop_config.json
- macOS:~/Library/Application Support/Claude/claude_desktop_config.json

{ "mcpServers": { "sql-server": { "command": "python", "args": ["-m", "sql_server_mcp.server"], "cwd": "C:\\Users\\benha\\Desktop\\sql-server-mcp", "env": { "SQL_SERVER_HOST": "192.168.1.117", "SQL_SERVER_DATABASE": "EM_Data", "SQL_SERVER_USERNAME": "benhg", "SQL_SERVER_PASSWORD": "your_password_here", "SQL_SERVER_PORT": "1433" } } } }

- Install the "Model Context Protocol" extension
- Add to VS Code settings.json:

{ "mcp.servers": { "sql-server": { "command": "python", "args": ["-m", "sql_server_mcp.server"], "cwd": "C:\\Users\\benha\\Desktop\\sql-server-mcp" } } }

{ "mcp": { "servers": { "sql-server": { "command": "python", "args": ["-m", "sql_server_mcp.server"], "cwd": "C:\\Users\\benha\\Desktop\\sql-server-mcp" } } } }

- execute_query- Run SQL queries with safety limits
- get_schema- Inspect database structure
- get_table_info- Detailed table information with samples
- explain_query- Query execution plans
- check_connection- Database connectivity status
- get_table_stats- Table size and performance metrics
- search_tables- Find tables and columns by name
- backup_table- Create table backups
- insert_data- Insert data with conflict handling

Once integrated with Claude, you can ask:

- "Show me the schema of my SeriesRecord table"
- "Execute: SELECT TOP 10 * FROM SeriesRecord WHERE Source = 'NBP'"
- "What tables do I have in my database?"
- "Create a backup of my SeriesRecord table"
- "Search for any tables containing 'GDP'"
- "Show me statistics for all my tables"

- Automatic query limits (TOP 1000 by default)
- Parameterized query support
- Environment-based configuration
- Connection pooling and health checks
- Comprehensive error handling

pip install pytest pytest-asyncio pytest tests/

sql-server-mcp/ ├── sql_server_mcp/ │ ├── __init__.py │ └── server.py ├── tests/ │ ├── __init__.py │ └── test_server.py ├── requirements.txt ├── pyproject.toml ├── .env └── README.md

- ODBC Driver Not Found: Install Microsoft ODBC Driver 17 for SQL Server
- Connection Failed: Verify server address, credentials, and network connectivity
- Permission Denied: Ensure database user has appropriate permissions

# Test ODBC drivers import pyodbc print(pyodbc.drivers()) # Test basic connection from sqlalchemy import create_engine, text engine = create_engine('mssql+pyodbc://user@server:port/database?driver=ODBC+Driver+17+for+SQL+Server') with engine.connect() as conn: result = conn.execute(text('SELECT 1')) print('Connection successful!')

MIT License - see LICENSE file for details.

Multi-database agent access (PostgreSQL, SQLite, MySQL, Oracle, SQL Server) with batch queries, pre-configured connections, and SQLGlot-enforced read-only safety

MCP servers that give AI clients safe, structured access to SQL databases. One installable package per database engine. Every package exposes the same minimal two-tool surface, execute_sql and search_objects, with guardrails on by default: read-only mode, row caps, and statement timeouts.

Guardrails for natural-language database queries.

Provides read-only access to Apache Iceberg tables via Apache Impala, allowing LLMs to inspect schemas and execute queries.

MCP server that gives AI agents deep understanding of contemporary or legacy relational databases.

A bridge for AI assistants to directly query and explore Microsoft SQL Server databases.

Production safe MCP server that gives AI agents direct access to MySQL databases.

MCP server for MySQL/MariaDB — schema inspection, queries, and data manipulation for AI assistants

Provides AI agents with direct access to query, search, and analyze MySQL databases.

A read-only MySQL database server for LLMs to inspect schemas and execute queries.