diogenes
Pass
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional and persona-based. It contains no executable scripts, command line operations, or network requests.
- [SAFE]: No obfuscation, hidden content, or suspicious encoding techniques were found in either the primary skill file or the reference documentation.
- [SAFE]: The tool access is limited to 'Read' and 'AskUserQuestion', which is appropriate for a dialogue-based persona and presents a minimal attack surface.
- [SAFE]: All references to external files (like '../../framework.md') are local to the assumed environment and follow standard modular configuration patterns.
- [SAFE]: The instructions for the AI persona ('challenge through absurdity', 'expose pretension') are consistent with the historical figure of Diogenes and do not represent attempts to bypass safety filters or ignore system instructions.
Audit Metadata