This is Anthropic's official approach to systematically profiling datasets before you dive into analysis. It walks you through structural understanding (row counts, grain, primary keys), column-level profiling (types, nulls, distributions), and pattern discovery in a methodical way. Think of it as a checklist that prevents you from making assumptions about data you haven't actually looked at yet. With 233 installs and three security audit passes, it's clearly being used in production workflows. The methodology is opinionated but comprehensive, which is exactly what you want when you're staring at an unfamiliar CSV or database table and need to know where to start.
npx skills add https://github.com/anthropics/knowledge-work-plugins --skill data-exploration