etsy-tony-scene-photo-3
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided text descriptions and product images to generate image prompts, which represents an attack surface for indirect prompt injection. 1. Ingestion points: User-provided product reference photos and text descriptions (as seen in SKILL.md). 2. Boundary markers: There are no explicit delimiters or instructions provided to separate user-provided content from the system prompt during interpolation. 3. Capability inventory: The skill has the ability to call image generation tools and write files to the local directory 'etsy-output/scene-photo-3'. 4. Sanitization: The skill instructions focus on visual verification but do not mention sanitizing or filtering user input before use in tool prompts.
Audit Metadata