Data-Forge

概要

Data-Forge functions as a Model Context Protocol (MCP) server, designed to empower any Large Language Model with extensive data science capabilities. It offers a comprehensive suite of high-performance tools covering the entire data lifecycle, from efficient data loading, cleaning, and validation to in-depth profiling, advanced feature engineering, and sophisticated visualization. This robust toolkit enables LLMs to effectively manage, analyze, and interpret complex datasets, making them invaluable for data-driven insights and exploration.

主な機能

• Load CSV or Parquet files and manage multiple datasets using Pandas or Polars.
• Validate dataset schemas with Pandera and clean data using PyJanitor pipelines.
• Generate comprehensive statistical profiles and insights with YData Profiling.
• Create and save various Seaborn/Matplotlib charts for LLM interpretation.
• 0 GitHub stars
• Execute complex SQL queries across loaded datasets using an in-memory DuckDB engine.

ユースケース

• Clean, validate, and visualize datasets directly through LLM commands and receive image outputs.
• Automate data loading and comprehensive profiling for immediate insights via LLM actions.
• Acquire and process data by scraping structured tables from web URLs or loading from Hugging Face Hub.