head-pop-out
Warn
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the user to install an external dependency named 'video-effects' and several heavy components (ffmpeg, Node, Puppeteer, torch). The source and integrity of this engine are not specified within the skill metadata.
- [COMMAND_EXECUTION]: The instructions direct the user to execute a JavaScript file located at a specific local path ('~/.claude/skills/video-effects/run.cjs') using Node.js. This represents execution of code outside the immediate skill package.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which could contain malicious instructions.
- Ingestion points: The skill ingests a user-provided video file ('my-clip.mp4') and a text string via the '--words' argument.
- Boundary markers: No boundary markers or 'ignore' instructions are present to delimit user-provided text from the prompt context.
- Capability inventory: The skill utilizes significant system capabilities including Node.js execution, ffmpeg for media processing, and Puppeteer for browser automation.
- Sanitization: There is no evidence in the SKILL.md file of input validation, escaping, or sanitization for the '--words' parameter or video metadata.
Audit Metadata