DuckDB

Provides direct access to DuckDB's analytical capabilities through file loading, table management, schema inspection, and SQL query execution for fast local processing of CSV, JSON, and Parquet files without database setup overhead.

MCPlucker is a Go-based MCP server that provides direct access to DuckDB's analytical capabilities through five core tools: file-to-table loading (CSV, JSON, Parquet), table management, schema inspection, and SQL query execution with configurable result limits. Built with the go-duckdb driver and MCP Go SDK, it uses an in-memory DuckDB instance by default and includes comprehensive error handling, automatic file type detection, and structured JSON responses for all operations. The implementation follows a clean architecture with separate packages for database management, server logic, and tool implementations, making it ideal for data analysts and developers who need fast, local analytical processing of structured data files without the overhead of a full database setup.