The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted data providing a surface for indirect prompt injection. * Ingestion points: Audio and video content is transcribed into a .srt file as described in SKILL.md. * Boundary markers: There are no explicit delimiters or instructions to ignore commands within the transcribed text during the cleanup pass. * Capability inventory: The skill instructions include the use of curl, sed, rm, and yt-dlp. * Sanitization: No filtering or validation is performed on the transcribed text before it is reviewed by the agent for errors.
[EXTERNAL_DOWNLOADS]: The skill fetches machine learning models from Hugging Face's official repository.
[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands for media processing (ffmpeg, ffprobe), downloading content (yt-dlp), and text manipulation (sed, grep).

video-subtitler