image-to-code
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill defines a strictly instructional workflow for web design and implementation. It does not utilize network tools, access sensitive files, or execute external code.
- [INDIRECT_PROMPT_INJECTION]: The skill involves extracting text from AI-generated images to be used in frontend implementation (Rule 21). This represents a theoretical attack surface for indirect prompt injection.
- Ingestion points: User prompts and text extracted from generated images (Rule 21).
- Boundary markers: The skill provides a structured set of rules (1-38) but does not define explicit delimiters or sanitization for extracted text.
- Capability inventory: File implementation (Rule 26).
- Sanitization: No explicit sanitization or validation of extracted text is described.
Audit Metadata