higgsfield-fight-scenes
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill performs legitimate prompt engineering for video generation with no detected malicious patterns.
- [EXTERNAL_DOWNLOADS]: The skill does not perform any external downloads. It relies on internal reference files and standard Higgsfield MCP tools.
- [COMMAND_EXECUTION]: No instances of shell command execution, subprocess spawning, or unauthorized system access were found.
- [DATA_EXFILTRATION]: The skill does not access sensitive local files, environment variables, or credentials. There are no network operations to non-whitelisted domains.
- [PROMPT_INJECTION]: The instructions focus on guiding the user through a creative process and do not contain patterns for bypassing safety guidelines or overriding agent behavior.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user input to craft prompts but includes a mandatory 'Present for review' step. This human-in-the-loop checkpoint ensures the user sees and approves the final prompt before any tools are invoked, mitigating the risk of accidental execution of injected content.
Audit Metadata