The Agent Skills Directory

[SAFE]: The skill implements video processing logic by invoking the ffmpeg and ffprobe system utilities. Commands are constructed as lists and executed via subprocess.run without a shell, which prevents command injection vulnerabilities.
[SAFE]: Dependency management uses uv with standard, reputable packages (typer, rich). No suspicious or unversioned external dependencies are loaded at runtime.
[SAFE]: No obfuscation, data exfiltration, or persistence techniques were detected. The scripts handle file paths and metadata securely, following standard Python practices.
[SAFE]: The skill includes a bundled font file for watermarking, which is a common practice to ensure consistent rendering across different systems and does not pose a security risk.

media-use