alterlab-bioservices

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

• Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md and included scripts (e.g., protein_analysis_workflow.py, pathway_analysis.py, compound_cross_reference.py, batch_id_converter.py) explicitly fetch and parse data from public third‑party resources (UniProt, KEGG, ChEMBL, ChEBI, UniChem, PSICQUIC, NCBI BLAST, QuickGO, etc.), and those external entries are read and used to drive mappings, queries, and follow‑up actions, so untrusted third‑party content can materially influence the agent's behavior.