The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Python subprocess module to execute external binaries including ffmpeg, ffprobe, and yt-dlp for video and metadata processing. These calls use argument lists rather than shell strings, minimizing command injection risks.
[COMMAND_EXECUTION]: The setup.py script includes a function to automatically install missing dependencies on macOS using the Homebrew (brew) package manager if they are not found on the system path.
[EXTERNAL_DOWNLOADS]: The skill utilizes the yt-dlp tool to download video files and subtitle data from external URLs provided by the user (e.g., YouTube, Vimeo, TikTok).
[EXTERNAL_DOWNLOADS]: Audio data is extracted and transmitted to external well-known services (Groq or OpenAI) for transcription via their respective APIs when local captions are unavailable.
[SAFE]: The skill implements security best practices for credential management, storing API keys in a local configuration file (~/.config/watch/.env) and explicitly applying restrictive file permissions (chmod 600) to prevent unauthorized access by other local users.
[PROMPT_INJECTION]: The skill ingests untrusted data in the form of video transcripts and on-screen text. While this creates an indirect prompt injection surface, the impact is low as the data is used for summarization, and the skill includes basic sanitization to strip formatting tags.

watch-video