sound-design
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from a
timeline.jsonfile to drive the sound synthesis process. - Ingestion points:
timeline.json(referenced inSKILL.md). - Boundary markers: None mentioned in the instructions for isolating external data.
- Capability inventory: The skill utilizes shell commands for Python (
python3), Node.js (node), and package management (npm run) to render and verify audio. - Sanitization: No sanitization or validation steps for the input JSON are described.
- [EXTERNAL_DOWNLOADS]: The documentation explicitly instructs the user to clone an external toolkit from a GitHub repository (
https://github.com/whaleyxbt/claude-motion). This repository provides the necessary environment and scripts referenced throughout the skill. - [COMMAND_EXECUTION]: The skill provides various shell commands for the agent to execute, such as rendering audio cues via
python3 -m sfx sfx/cues/<video>.py, generating waveforms withnpm run wave, and performing loudness checks withnode scripts/wave.mjs.
Audit Metadata