patsnap-lifescience-target-intelligence
Pass
Audited by Gen Agent Trust Hub on Apr 22, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests and processes data from external sources such as patents, literature, and web search results. However, the skill's capabilities are limited to data retrieval and structured reporting, and it does not possess dangerous execution tools, making the risk negligible.
- [DATA_EXFILTRATION]: Uses domain-specific search and fetch tools for biological and clinical data. No access to sensitive local files, environment variables, or unauthorized network exfiltration was found.
- [SAFE]: The execution principles and analysis paths are consistent with the vendor's (patsnap) stated purpose. No obfuscation, persistence mechanisms, or privilege escalation patterns were identified.
Audit Metadata