The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of subprocess.run and subprocess.Popen to invoke ffmpeg and ffprobe. These operations are strictly scoped to media decoding, encoding, and feature extraction, which constitutes the primary intended purpose of the skill.
[EXTERNAL_DOWNLOADS]: The skill includes integration with the ElevenLabs API (api.elevenlabs.io) for generating text-to-speech audio clips. ElevenLabs is a well-known service, and the network operations are limited to data exchange for voice synthesis.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, including video files, audio tracks, and SRT subtitle files. While this creates an attack surface for indirect prompt injection, the skill's processing logic is focused on mathematical transformations and imaging rasterization rather than natural language interpretation, significantly mitigating the risk.
[COMMAND_EXECUTION]: A cleanup function (cleanup_render_artifacts) is provided to remove intermediate render files. It uses shutil.rmtree and os.remove to delete temporary directories and files matching specific patterns (e.g., *.wav, ffmpeg_*.log) created during the video production process.

ascii-video