video-poster-design
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project data (user briefs, character descriptions, and storyboard metadata) and interpolates them into prompts for an image generation tool. While this presents a theoretical injection surface, it is a functional requirement for the skill's purpose.
- Ingestion points: Processes content from user inputs,
characters/folder, andstoryboard/folder. - Boundary markers: Uses a structured prompt template with clear sections for context, composition, and constraints.
- Capability inventory: Accesses the
image_gentool and performs file write operations within the project-localposters/directory. - Sanitization: Relies on the underlying image generation model's safety filters and specific negative prompt constraints included in the template.
Audit Metadata