paper-collage-explainer-generator
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-supplied narration, story beats, or core concepts to generate media prompts.
- Ingestion points: User input is ingested at Step 1 (Parse the Input) in both
SKILL.mdandSKILL.cn.mdto extract core meaning and visual metaphors. - Boundary markers: The instructions do not specify explicit delimiters (such as XML tags or unique markers) to isolate user-provided data from the agent's internal instruction templates during interpolation.
- Capability inventory: The skill utilizes image generation, video generation (specifically the
MiniMax-H3model), and audio synthesis tools. - Sanitization: The skill employs a semantic translation step where it converts raw user input into a structured 'Production Plan' and 'Visual Metaphors', which serves as a natural sanitization layer before final prompt generation.
Audit Metadata