You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Add a comprehensive Model Context Protocol (MCP) server for Apache
Cloudberry enabling seamless integration with LLM applications.
This implementation provides secure database interaction capabilities
through AI-ready interfaces, supporting both stdio and HTTP transport
modes for maximum compatibility with various LLM clients.
**What:**
- Complete MCP server implementation with async database operations
- Safe SQL query execution with parameterized queries
- Administrative tools for performance monitoring and optimization
- Predefined prompts for common database tasks
- Security-first design with SQL injection prevention
**Why:**
- Enable AI assistants to interact with Cloudberry databases safely
- Provide standardized interface for LLM applications
- Reduce manual database management overhead
- Support modern AI development workflows
**How:**
- Built with asyncpg for high-performance PostgreSQL operations
- Implements MCP protocol for universal LLM compatibility
- Uses environment-based configuration for flexible deployment
- Includes comprehensive test suite and documentation
**Compatibility:**
- Requires Python 3.8+ with asyncpg support
- Compatible with Claude Desktop, Cursor, Windsurf, VS Code, Trae, Qwen Desktop, etc.
- Supports both stdio and HTTP transport modes
- Works with standard PostgreSQL connection strings
**Breaking Changes:**
None - this is a new addition.
0 commit comments