The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes ffmpeg and ffprobe to perform video processing tasks such as frame extraction, blending, and encoding. These commands are invoked using list-based arguments which prevents shell injection.
[EXTERNAL_DOWNLOADS]: The Python script uses torch.hub to download a legitimate open-source background removal model (RobustVideoMatting) from GitHub. It also requires several standard Python and Node.js libraries for its operations.
[DYNAMIC_EXECUTION]: A static analysis warning regarding an eval() call in prep.py was identified as a false positive. The code uses the PyTorch model.eval() method to set the neural network to evaluation mode for inference, which is safe and unrelated to Python's eval() function.
[SAFE]: The skill implements security best practices, such as escaping user-provided text in the engine.js file before rendering it in the browser to prevent potential cross-site scripting (XSS) in the video generation pipeline.

video-effects