This is a structured framework for post-market drug safety analysis using FAERS adverse event reports, FDA labels, and pharmacogenomic data. It pushes you to quantify signals with disproportionality measures like PRR and ROR rather than just describing risks, and includes clinical reasoning strategies like on-target versus off-target thinking and timeline-based mechanism classification. The compute-first approach means you write Python to calculate statistics rather than theorizing. Good for investigating safety signals, comparing drug profiles, or building pharmacovigilance reports when you need actual numbers and not just label summaries. The pharmacogenomic integration helps flag genetic risk factors for specific populations.
npx skills add https://github.com/mims-harvard/tooluniverse --skill tooluniverse-pharmacovigilance