meta-muse-video-analysis

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local Python script scripts/analyze_video.py to perform the video analysis. This is the primary intended function of the skill and uses standard argument passing for the video path and user prompt.
  • [EXTERNAL_DOWNLOADS]: The skill communicates with https://api.meta.ai/v1 to upload video files and retrieve analysis results. This is an official API endpoint for Meta's AI services and is documented as the intended backend for this skill.
  • [CREDENTIALS_UNSAFE]: The skill manages the Meta API key by reading it from the META_MUSE_KEY environment variable. The instructions and the script explicitly include logic to prevent the key from being printed, logged, or exposed in command-line arguments or agent outputs.
  • [DATA_EXFILTRATION]: The skill involves uploading local video files to Meta's servers. The skill instructions include clear warnings that data is being shared externally and require the script to delete the remote copy immediately after the analysis is complete unless the user overrides this behavior.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 02:22 PM
Security Audit — agent-trust-hub — meta-muse-video-analysis