varg-character-fitness
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill explicitly instructs the agent to use 'catalog-style wording' designed to get content 'past false moderation blocks' on image generation platforms.
- [PROMPT_INJECTION]: Contains a dedicated substitution table specifically for circumventing safety guards, providing synonyms for prohibited or high-risk terms (e.g., replacing 'sweaty' with 'light post-workout sheen' and 'posing' with 'mid-set') to ensure the resulting prompts bypass automated filters.
Audit Metadata