30-thiet-ke-master
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a well-documented design tool intended for creating marketing and branding assets using AI image generation models.
- [COMMAND_EXECUTION]: The skill utilizes conditional logic to verify environment configurations, such as checking for an OpenAI API key or an enterprise-specific binary (
OD_BIN). These checks are used solely for tier detection and selecting the appropriate generation route. - [DATA_EXPOSURE]: The skill searches the project directory for brand-related files (e.g.,
logo.svg,brand-guideline.md,prd.md) to extract visual styles and colors. This data access is essential for its function as a brand-aware design assistant. - [INDIRECT_PROMPT_INJECTION]: The skill accepts user-provided reference images and stylistic instructions. It mitigates potential risks by using a structured prompt architecture and explicit 'negative prompts' to ensure the output adheres to safety and quality guidelines.
Audit Metadata