mockup-device-3d
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and render user-provided text or HTML content within simulated device screens. This creates a surface for processing untrusted data, although the intended use is for visual mockup generation. 1. Ingestion points: User-provided data is interpolated into the screen div elements in the generated HTML. 2. Boundary markers: None explicitly defined in the templates. 3. Capability inventory: Limited to HTML/CSS rendering within the agent's UI or browser tool. 4. Sanitization: No specific sanitization logic is provided in the instructions.
- [EXTERNAL_DOWNLOADS]: The implementation fetches the Tailwind CSS framework from a well-known CDN (tailwindcss.com) and typography from Google Fonts. It also references resources from a well-known AI service (HeyGen).
Audit Metadata