adjust-logos
Pass
Audited by Gen Agent Trust Hub on Jul 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill content is purely documentation and CSS examples for design workflows. It contains no executable scripts, commands, or network operations.
- [PROMPT_INJECTION]: The skill includes behavioral constraints for the agent (e.g., "Never tokenize optical nudges"). These are task-specific functional requirements for design integrity rather than attempts to bypass AI safety filters or extract system instructions.
- [DATA_EXFILTRATION]: No sensitive file paths, credential access, or external network requests were identified.
- [REMOTE_CODE_EXECUTION]: There are no patterns involving package installation or the execution of remote scripts.
Audit Metadata