This is a comprehensive drug target intelligence gathering workflow that pulls data from nine parallel research paths including tissue expression (GTEx, HPA), protein interactions (STRING), genetic variants (ClinVar, gnomAD), and druggability databases (DGIdb, ChEMBL). It's built for answering "tell me everything about target X" queries in drug discovery, with mandatory evidence grading (T1-T4), inline citations, and a report-first approach that creates a structured markdown file before populating it progressively. The workflow includes a target evaluation framework that walks through genetic evidence, druggability, safety, and competitive landscape. The instructions are extremely prescriptive about avoiding hallucinations (look up, don't guess) and handling tool parameter verification before making calls. Worth noting it explicitly documents negative results rather than leaving empty sections, which is good practice for research transparency.
npx skills add https://github.com/mims-harvard/tooluniverse --skill tooluniverse-target-research