The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes system-level commands to process media files and generate videos.
Evidence: scripts/video_maker.py and scripts/verify_alignment.py utilize the subprocess.run module to call ffmpeg and ffprobe.
Context: Commands are constructed using argument lists rather than shell strings, which is a standard security practice to prevent shell injection vulnerabilities.
[EXTERNAL_DOWNLOADS]: The skill requires several external tools and libraries to function.
Evidence: The README.md and SKILL.md specify dependencies on ffmpeg, edge-tts, and pyyaml.
Context: These are widely recognized and trusted open-source utilities and libraries. The edge-tts package communicates with official Microsoft Edge TTS services, which is the intended purpose of the skill.
[DATA_EXPOSURE]: The skill processes project-specific configuration and assets.
Evidence: Scripts read narration.json, video_config.yaml, and user-generated images/audio within the assets/generated/ workspace.
Context: File access is confined to the local project environment, with no evidence of accessing sensitive system paths (e.g., SSH keys) or unauthorized network exfiltration.

video-creator