claude-video-analyze

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides numerous Bash command recipes for ffprobe and ffmpeg to inspect video properties. It includes a multi-step script for testing Constant Rate Factor (CRF) values, which writes temporary video files and results to /tmp/ before removing them.
  • [PROMPT_INJECTION]: The skill contains a directive stating that operations are "safe to auto-execute without confirmation," which encourages the agent to skip standard user verification for shell commands. Additionally, the skill demonstrates a surface for indirect prompt injection via user-provided filenames.
  • Ingestion points: User-provided file paths assigned to the $INPUT variable in SKILL.md.
  • Boundary markers: None identified.
  • Capability inventory: Uses the Bash tool for shell execution.
  • Sanitization: No explicit escaping or validation of the $INPUT variable is present.
  • [SAFE]: The external utilities referenced (ffmpeg, ffprobe, scenedetect) are standard industry tools for video processing, and their usage here aligns with the skill's stated purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 02:55 PM
Security Audit — agent-trust-hub — claude-video-analyze