disease-mechanism-evidence-map
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions and associated reference modules focus exclusively on structured medical evidence synthesis.
- [SAFE]: Hard rules 11-14 provide robust safeguards against the fabrication of scientific literature, DOIs, and PMIDs.
- [SAFE]: No dangerous capabilities such as network access, file system modification, or arbitrary command execution were identified in the instruction set.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided research topics to generate mechanism maps; however, the lack of exploitable tools or system-level access renders this attack surface benign.
- [SAFE]: The skill author (AIPOCH) uses legitimate vendor resources for repository links and consistent metadata.
Audit Metadata