The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local shell commands to facilitate its core automation features.
The scripts/check_video.py script invokes ffprobe via subprocess.run to validate video metadata against platform constraints.
The scripts/notify_user.py script calls system binaries including osascript, afplay, and say using subprocess.run and subprocess.Popen for macOS notifications and text-to-speech functionality.
[INDIRECT_PROMPT_INJECTION]: The skill maintains an attack surface for indirect prompt injection due to its interaction with and ingestion of content from external publishing platforms.
Ingestion points: Text content such as video titles, descriptions, and user-edited tags are read from the DOM of WeChat Channels and Bilibili pages using snapshotText() and browser automation.
Boundary markers: The instructions do not employ specific boundary markers or delimiters when processing content retrieved from platform pages.
Capability inventory: The skill has access to command execution through its included Python scripts and wide-ranging browser automation capabilities.
Sanitization: There is no explicit mechanism described for sanitizing or validating text read back from external websites before it is used to inform further agent actions.

lov-media-publisher