seedance-vocab-es
Warn
Audited by Gen Agent Trust Hub on Apr 23, 2026
Risk Level: MEDIUMPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill contains a section titled 'Production Context Tokens (Filter Shield)' (Section E.15) which provides phrases intended to frame potentially restricted content in a way that bypasses automated safety filters of generation models.\n- [PROMPT_INJECTION]: Sections E.7 ('Safe Action & Combat Vocabulary') and E.8 ('Safe Weapon Vocabulary') provide euphemistic and descriptive alternatives for potentially prohibited terms (e.g., using 'instrumento de acero alargado' instead of 'espada' or 'sword'), which are explicit techniques for circumventing content moderation guidelines.
Audit Metadata