embedded-captions

Pass

Audited by Gen Agent Trust Hub on Jul 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the download of AI models and tools necessary for its operation. This includes fetching the WhisperX transcription engine via uvx and human segmentation models (u2net_human_seg) for subject matting. These are standard functional dependencies used to provide the skill's core AI capabilities.
  • [COMMAND_EXECUTION]: The skill uses shell scripts and Node.js subprocesses to execute media processing commands. These involve FFmpeg for video rendering, ffprobe for metadata extraction, and Puppeteer for capturing pixel-perfect text layouts. These executions are scoped to project-specific media and align with the tool's primary purpose of video production.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 21, 2026, 11:08 PM
Security Audit — agent-trust-hub — embedded-captions