This matches cancer patients to clinical trials by their molecular profile, not just diagnosis. It queries ClinicalTrials.gov then cross-references against CIViC, OpenTargets, ChEMBL, and FDA databases to score trials on a 0-100 scale with evidence tiers. The workflow is thorough: it parses eligibility criteria for biomarker requirements, maps patient mutations to trial interventions, checks drug mechanisms against the patient's specific alterations, and handles geographic proximity. Best for precision oncology cases where you need ranked trial options for patients with known mutations like EGFR L858R or KRAS G12C, especially after standard therapies have failed. The molecular-first matching is the key bit here, it prioritizes genotype over everything else.
npx skills add https://github.com/mims-harvard/tooluniverse --skill tooluniverse-clinical-trial-matching