media-higgsfield-video
Pass
Audited by Gen Agent Trust Hub on Aug 19, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of instructional markdown files and configuration guidelines. It does not include any executable code, scripts, or binary files.
- [SAFE]: The skill implements a robust 'live lookup' design. Instead of hardcoding model IDs or parameters, it instructs the agent to use the
models_exploretool to fetch real-time constraints and schemas from the server, preventing configuration drift and errors. - [SAFE]: All external URLs referenced in the documentation point to official vendor sites (e.g., Google, Alibaba, ByteDance, xAI) or the vendor's own repositories (Higgsfield), which are trusted or well-known technology services.
- [SAFE]: The skill explicitly includes safety constraints, directing the agent not to generate NSFW content or content that violates copyright or portrait rights.
Audit Metadata