bioinformatics

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

• Third-party content exposure detected (high risk: 0.80). The skill's Step-by-Step methodology and "Key Databases and Tools" explicitly require fetching and using data from public third-party databases (e.g., STRING, KEGG, UniProt, NCBI) so the agent will ingest and act on external, untrusted web content that can materially influence analysis and subsequent tool use (e.g., network construction, enrichment calls).