ai-elements
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches media assets (audio, images) and web preview content from well-known and trusted services, including Vercel (vercel-storage.com, vusercontent.net), Unsplash (images.unsplash.com), and official documentation sites (react.dev).
- [DYNAMIC_EXECUTION]: The JSXPreview component is designed to dynamically render JSX strings generated by AI models. While this involves dynamic rendering, it is the primary intended functionality of the component for displaying generated UI in real-time.
- [INDIRECT_PROMPT_INJECTION]: The skill contains components like MessageResponse and JSXPreview that ingest and render external data (Markdown, JSX). While these provide a surface for indirect injection, they are standard implementations for AI chat interfaces and are considered low risk in this context.
Audit Metadata