What problem does it solve? Connecting AI assistants like Claude or Cursor to databases normally requires custom integration code for each database type. This Skill configures a universal MCP connector that lets AI tools query MySQL, PostgreSQL, MongoDB, Oracle, and 13 other databases using natural language, without writing boilerplate connection logic. ## Core Features & Use Cases - Multi-Database MCP Server: Supports 17 database types including MySQL, PostgreSQL, MongoDB, Redis, Oracle, SQL Server, SQLite, and Chinese databases like Dameng, KingbaseES, and GaussDB. - Dual Operation Modes: Runs in stdio mode for Claude Desktop and Cursor, or HTTP mode exposing MCP over SSE/Streamable HTTP plus a REST API for Dify and remote access. - Built-in MCP Tools: Provides execute_query, get_schema, get_table_info, get_sample_data, connect_database, and cache management tools, with read-only mode and automatic sensitive data masking. - Use Case: Configure Claude Desktop with a read-only connection to a production MySQL database, then ask questions like "What are the top 10 products by revenue this month?" and Claude automatically inspects the schema, generates SQL, and explains the results. ## Quick Start Set up the universal-db-mcp connector in Claude Desktop to connect to my local MySQL database in read-only mode so I can query it with natural language.