shortcut-defense
Pass
Audited by Gen Agent Trust Hub on May 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of markdown instructions and a JSON test file with no executable scripts (Python, Node.js, or Shell).
- [EXTERNAL_DOWNLOADS]: No external URLs or remote resources are referenced for download or execution.
- [DATA_EXFILTRATION]: There are no network operations or sensitive file system access patterns present.
- [PROMPT_INJECTION]: The instructions follow a clear psychological methodology and do not attempt to override agent safety protocols or extract system instructions.
- [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or credentials were detected in any of the files.
Audit Metadata