feynman-perspective
Pass
Audited by Gen Agent Trust Hub on May 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a cognitive framework and a specific persona based on historical and academic sources without attempting to override agent safety protocols.
- [SAFE]: No prompt injection techniques or markers, such as instructions to ignore previous rules or bypass filters, were found in the text.
- [NO_CODE]: The skill is entirely descriptive and does not include any executable scripts, binaries, or instructions for installing external software dependencies.
- [EXTERNAL_DOWNLOADS]: While the instructions mention using tools like WebSearch to verify facts, this represents a standard use of agent capabilities for research and does not constitute a remote code execution risk.
- [DATA_EXFILTRATION]: There are no patterns indicating attempts to access sensitive file paths, harvest credentials, or exfiltrate private user information to external servers.
Audit Metadata