higgsfield-recipes

Pass

Audited by Gen Agent Trust Hub on Oct 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely informational, consisting of markdown templates and instructional text for guiding video generation prompts. It does not include any scripts, executable code, or commands that interact with the host system.
  • [INDIRECT_PROMPT_INJECTION]: The skill uses a template-based approach where user-supplied descriptions are interpolated into pre-defined prompt structures (e.g., [Subject description], [Obstacle]). This represents an indirect prompt injection surface; however, as the skill lacks execution capabilities, network access, or sensitive file system tools, the risk is minimal. The documentation also directs users to reference safety constraints in negative-constraints.md.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 26, 2026, 11:52 AM