anime-animation-production
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a process for the agent to ingest and act upon untrusted user creative requests, which constitutes an indirect prompt injection surface. \n- Ingestion points: Ingests user-provided requests for anime production, including character descriptions, genre styling, and shot plans as detailed in SKILL.md. \n- Boundary markers: The skill provides instructions for the agent to re-assert core character and style tags in every prompt and to anchor generations to a fixed model sheet to prevent identity drift. \n- Capability inventory: The instructions direct the agent to utilize external generative media tools (e.g., Sora, Kling, Vidu, Midjourney) and to perform post-processing tasks like frame decimation. \n- Sanitization: Part 7 contains explicit instructions for the agent to identify and refuse requests involving the imitation of living artists or reproduction of protected IP, offering original creative alternatives instead.
Audit Metadata