sound-design-and-music
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized access to sensitive data were found. The skill contains legitimate instructional content for audio production workflows.
- [INDIRECT_PROMPT_INJECTION]: The skill processes film treatments and shot lists which could potentially contain malicious instructions intended to influence the generated audio prompts.
- Ingestion points: Data is ingested from the
sound.ambienceandsound.musicfields of a film treatment and the project's shot list (SKILL.md). - Boundary markers: Absent. The skill relies on natural language instructions to separate input data from prompt generation logic.
- Capability inventory: The skill utilizes local scripts (
scripts/sfx.py) and ComfyUI workflows (generateMusic) for asset generation. - Sanitization: The skill mandates manual checks for named artists or films and uses negative constraints in instructions (e.g., "no music" in effects prompts) to prevent unwanted model behavior.
Audit Metadata