The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute yt-dlp and ffmpeg commands on the host system to download and process media files. This is the primary intended function of the skill.
[EXTERNAL_DOWNLOADS]: The skill recommends installing dependencies (yt-dlp, ffmpeg) via official package managers such as PyPI, Homebrew, or APT. It also suggests using the tool's built-in update command (yt-dlp -U).
[INDIRECT_PROMPT_INJECTION]: The skill defines a vulnerability surface where untrusted data (URLs) is passed to command-line tools, and the resulting output is processed by the agent.
Ingestion points: External URLs provided to the yt-dlp command and JSON metadata fetched from those URLs via the --dump-json flag (SKILL.md).
Boundary markers: Absent; the skill relies on the agent correctly quoting the URL in the shell command to prevent command injection.
Capability inventory: Shell command execution (yt-dlp, ffmpeg) and file system write access for saving downloads.
Sanitization: Absent; there is no explicit validation or sanitization of the URL or the metadata returned from the network before it enters the agent's context.

video-download