datapitfalls (GitHub Repo) (opens in new tab)

datapitfalls is an open-source Claude-powered tool that audits charts, code, written analysis, and documents for common data mistakes across the full reasoning chain, from question framing to communication. It goes beyond chart linting by using Ben Jones' Avoiding Data Pitfalls taxonomy to identify issues like bias, silent pipeline failures, bad aggregation, statistical errors, misleading visuals, and unclear presentation, then explains severity and fixes.