The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute the yt-dlp command-line utility to fetch video titles, view counts, and auto-generated captions.
[COMMAND_EXECUTION]: The skill invokes local Python scripts located in sibling directories, specifically ../yt-package/scripts/title.py and ../yt-script/scripts/hookscore.py, to analyze video content.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data retrieved from YouTube video titles, descriptions, and auto-generated captions, which could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: Video metadata retrieved via list_channel_posts and captions fetched via yt-dlp in SKILL.md.
Boundary markers: None identified. Untrusted content is processed directly by analysis scripts.
Capability inventory: The skill has the ability to execute shell commands (yt-dlp) and local scripts via the shell.
Sanitization: No evidence of sanitization or filtering of video titles or captions before they are passed to analysis scripts.
[DATA_EXPOSURE]: The skill accesses a configuration file located at ~/.uplika/<workspace-slug>/voice.md to retrieve user-specific voice profiles. This is a vendor-owned resource associated with the 'uplika' ecosystem.

yt-audit