tooluniverse-variant-interpretation

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). This skill explicitly fetches and ingests content from open public databases and user-generated/preprint sources (e.g., PubMed/EuropePMC/BioRxiv/MedRxiv, ClinVar, gnomAD, COSMIC, PDB/AlphaFold, etc., as described in SKILL.md Phases 2–5 and the Checklist) and directly uses that untrusted third‑party content to drive ACMG evidence assignment, classifications, and clinical recommendations, so such content could materially influence agent actions.