tooluniverse-microbiome-research

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

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime workflow calls public third-party APIs (e.g., MGnify/EuropePMC/ENA/OLS/GTDB) whose responses include outsider-authored free text such as study abstracts and article titles/abstracts; this text is then ingested into the agent’s LLM context via the tool outputs (e.g., MGnify_get_study_detail, EuropePMC_search_articles, PubMed_search_articles, ENAPortal_search_studies).