Causal Inference

This helps Claude understand and apply causal inference methods when you're trying to figure out if something actually caused an outcome rather than just correlating with it. You'd reach for this when analyzing A/B test results, estimating policy impacts from observational data, or dealing with confounders in your analysis. It covers the standard toolkit: propensity scores, instrumental variables, and heterogeneous treatment effects. Honestly, this is one of those areas where having the right statistical framework matters a lot, because it's easy to make confident but wrong claims about causation. The skill passed security audits and sits in a repository with 245 stars, so it's getting some traction with folks doing data analysis work.