The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes system-level commands using yt-dlp and ffmpeg to download video segments and extract frames for analysis.
[EXTERNAL_DOWNLOADS]: The skill retrieves media files from YouTube URLs using the yt-dlp utility.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted transcript data from external YouTube videos and processes it using an LLM without safety boundaries.
Ingestion points: Video transcripts are fetched via the summarize tool as described in SKILL.md.
Boundary markers: None are specified; the transcript content is directly passed for analysis.
Capability inventory: The agent has access to yt-dlp for file downloads and ffmpeg for file processing.
Sanitization: There is no evidence of filtering, escaping, or instruction-aware sanitization applied to the ingested transcript data.

youtube-pro