tooluniverse-pharmacovigilance
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a structured workflow for medical research using specific, well-known scientific databases. All network activities are performed through defined tool calls to these services.
- [INDIRECT_PROMPT_INJECTION]: The skill has an inherent attack surface for indirect prompt injection because it ingests unstructured text from external sources like PubMed and BioRxiv. While these are reputable scientific repositories, the skill does not explicitly instruct the agent to use boundary markers or sanitize the ingested data before processing it. However, the capabilities are limited to report generation and scientific calculations, presenting a low risk.
- Ingestion points: Data retrieved from FAERS, DailyMed, PharmGKB, PubMed, BioRxiv, MedRxiv, and ClinicalTrials.gov.
- Boundary markers: None identified in the prompt instructions.
- Capability inventory: File creation (Markdown and CSV), internal tool execution for data retrieval.
- Sanitization: No explicit sanitization of tool outputs before interpolation into reports.
Audit Metadata