design-to-code
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided design images and screenshots, which represent an ingestion surface for untrusted data. Instructions hidden within images (e.g., via OCR) could attempt to influence the code generation process. The skill mitigates this by enforcing a three-phase workflow that requires user confirmation of the layout skeleton before proceeding to implementation.
- Ingestion points: User-provided design images, screenshots, and wireframes referenced in the skill instructions.
- Boundary markers: None explicitly defined in the prompts to distinguish between design content and potential embedded instructions.
- Capability inventory: The skill generates frontend code involving component usage, script logic for API calls, and routing.
- Sanitization: No specific sanitization or validation of image-extracted text is mentioned.
- [SAFE]: The skill references 'bkui-vue', which is a standard UI component library for the Vue framework, consistent with the vendor's development environment. No unauthorized remote code execution, persistence, or data exfiltration patterns were observed.
Audit Metadata