MCPQL (SQL Server)

This MCP server provides AI assistants with direct access to SQL Server databases through 10 specialized tools covering table analysis, stored procedure inspection, data preview with filtering, column statistics, query execution, and comprehensive database object search. Built by Hendrick Castro using TypeScript with the mssql library and Model Context Protocol SDK, it supports multiple authentication methods (SQL, Windows, Azure AD), connection pooling, and parameterized queries for security, organizing database functionality into categories for structural analysis (table schemas, indexes, constraints), code analysis (stored procedure parameters, dependencies, definitions), data operations (preview, statistics, sample values), and search capabilities (object names, definitions, dependencies). The implementation includes comprehensive error handling, parallel query execution for performance, and supports both local SQL Server instances and Azure SQL Database, making it valuable for database documentation workflows, data exploration tasks, and building AI assistants that need programmatic access to SQL Server metadata and data without manual database management tools.