The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on standard command-line tools like ffmpeg, ffprobe, file, and du to perform media processing. These commands are used according to their intended purpose and are safe within this context.
[PROMPT_INJECTION]: The skill processes untrusted input in the form of video files, which presents a surface for indirect prompt injection. * Ingestion points: External media files provided to the $INPUT_FILE variable in SKILL.md. * Boundary markers: None present to distinguish data from instructions for the agent. * Capability inventory: The skill executes shell commands (ffmpeg, ffprobe) on the input files. * Sanitization: The script uses double-quoting for variables (e.g., "$INPUT_FILE") which helps prevent command injection via malicious filenames.

gif-generation