tooluniverse-pharmacovigilance
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed for legitimate pharmacovigilance research, retrieving data from recognized medical and scientific databases such as DailyMed, ChEMBL, FAERS, and PharmGKB.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from sources like drug labels (DailyMed) and research abstracts (PubMed). While this presents a surface for indirect prompt injection, the risk is minimal as the skill's capabilities are restricted to generating markdown reports and structured CSV files. There are no high-risk capabilities like shell execution or network exfiltration of local system data.
- [COMMAND_EXECUTION]: The skill uses Python-like pseudo-code to guide the agent in using specific tools. These tools are part of a defined internal toolset (
tu.tools) for data retrieval and do not involve execution of arbitrary system commands.
Audit Metadata