Skill UI

This skill calculates and documents the residual re-identification risk of de-identified datasets. It goes beyond simple direct identifier removal by assessing the risk posed by quasi-identifiers (age, ZIP, dates). It computes structural metrics like k-anonymity and l-diversity, and runs empirical linkage attacks to provide a defensible memo supporting HIPAA Expert Determination for determining if a dataset is safe for release.