anesthesiology
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill content is restricted to academic and clinical publication guidelines. There are no instances of prompt injection, obfuscation, persistence mechanisms, or unauthorized privilege escalation.
- [INDIRECT_PROMPT_INJECTION]: The skill instructions identify a data ingestion surface as it directs the agent to read external resource files (../../resources/source-basis.md) and perform web searches for author instructions. However, no boundary markers or sanitization requirements are defined because the skill does not include exploitable capabilities such as arbitrary command execution or file system writing.
Audit Metadata