tooluniverse-polygenic-risk-score

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow in python_implementation.py calls gwas_get_associations_for_trait(...) using user-supplied trait and then ingests text fields from the returned association records (e.g., beta/or_value, p_value, mapped_genes, locations) into the PRS model without first selecting a specific trusted item.