tooluniverse-infectious-disease
Pass
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill aggregates large amounts of text from external, third-party sources like PubMed and bioRxiv, creating a surface for indirect prompt injection attacks.\n
- Ingestion points: Data is ingested from external literature and web searches via tools such as
PubMed_search_articles,EuropePMC_search_articles,ArXiv_search_papers, andweb_search.\n - Boundary markers: The agent instructions do not provide specific delimiters or ignore-behavior markers to prevent instructions within external text from influencing the agent.\n
- Capability inventory: The skill possesses capabilities to perform genome analysis, predict protein structures using NVIDIA NIM, and conduct molecular docking, providing a wide range of actions that could be influenced by injected instructions.\n
- Sanitization: No explicit sanitization or filtering logic is described for the text content retrieved from external sources before it is interpreted by the agent.\n- [EXTERNAL_DOWNLOADS]: The skill connects to multiple well-known and trusted scientific repositories and technical services.\n
- Evidence: The skill references and uses official APIs for NCBI (Taxonomy), UniProt (Proteins), ChEMBL (Drug Bioactivity), DrugBank, KEGG (Pathways), Reactome, NVIDIA (NIM), PubMed, and ClinicalTrials.gov for pathogen characterization and drug discovery.
Audit Metadata