video-loop

Pass

Audited by Gen Agent Trust Hub on Sep 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute ffmpeg and ffprobe for video metadata analysis and multi-format encoding (MP4 and WebM). The execution is limited to standard media processing utilities.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided media files, which represents an ingestion point for external data.
  • Ingestion points: The skill accepts a path to a video file via command-line arguments in SKILL.md and the video-loop.sh script.
  • Boundary markers: None; the skill relies on the inherent parsing limitations of FFmpeg and FFprobe for the specific media types.
  • Capability inventory: Capabilities are restricted to local file reading, system information retrieval (stat), directory creation (mkdir), and media processing via ffmpeg/ffprobe.
  • Sanitization: The provided bash script implements proper shell quoting for all file paths and variables, preventing common command injection vulnerabilities. Duration and size data are retrieved using tool-specific flags that output raw values.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 28, 2026, 05:32 AM