scene-distillation-zine-v1-3
Pass
Audited by Gen Agent Trust Hub on Aug 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill incorporates robust privacy instructions, explicitly directing the agent not to browse, search, share, or upload user-provided photos to external locations, nor save them to local project files.
- [SAFE]: The instructions establish clear boundaries for data processing, defining the user photo as a 'semantic reference only' and prohibiting the inclusion of photographic pixels in the final output, which mitigates risks associated with processing untrusted image data.
- [SAFE]: The agent is instructed to avoid displaying its generation prompt to the user, which is a common non-malicious technique used to protect instructions from leakage.
- [SAFE]: No unauthorized command execution, network exfiltration, or persistence mechanisms were found. The skill operates within the expected bounds of a creative prompt-engineering extension.
Audit Metadata