The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the yutu command-line interface to execute various playlist management tasks, including list, insert, update, and delete operations. These are carried out via shell commands within the agent's environment.
[EXTERNAL_DOWNLOADS]: The skill's setup guide facilitates the download and installation of the @eat-pray-ai/yutu utility from the npm registry and provides alternatives such as Homebrew, Winget, and direct binary downloads from GitHub.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it retrieves and processes external content from the YouTube API (e.g., video titles and descriptions) that could contain malicious instructions.
Ingestion points: Data entering the agent's context through the yutu playlistItem list command, as described in references/playlistItem-list.md and SKILL.md.
Boundary markers: The skill does not implement delimiters or provide instructions to the agent to disregard potentially malicious instructions embedded in the API response.
Capability inventory: The agent has access to the yutu CLI and other local environment tools to perform operations based on API data.
Sanitization: There is no evidence of sanitization or filtering applied to the content fetched from the YouTube Data API before it is processed by the agent.

youtube-playlist-item