The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes ffmpeg and ffprobe via Python's subprocess.run to render and validate video files. These calls use list-based arguments and resolve executable paths using shutil.which to minimize shell injection risks.
[DYNAMIC_EXECUTION]: render_video.py dynamically generates complex FFmpeg filter graphs and subtitle files (ASS format) based on the project's storyboard and configuration files.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function involves ingesting and processing data from external web applications and repository files, creating a potential surface for indirect instructions.
Ingestion points: Target web application UI (via Playwright automation) and repository documentation.
Boundary markers: The workflow enforces distinct phases (inspection, planning, validation, and rendering) with human-in-the-loop verification steps.
Capability inventory: Shell command execution for media processing, file system writes within isolated directories, and browser automation.
Sanitization: Includes path escaping for FFmpeg filters and mandates the use of environment variables for resolving sensitive authentication data.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of the playwright Node.js package from the official NPM registry.

web-product-demo-video