The Agent Skills Directory

[SAFE]: The skill operates entirely locally to process video files as described. No malicious behavior or suspicious patterns were identified during the analysis.
[COMMAND_EXECUTION]: The skill invokes external system utilities to perform media processing and hardware detection.
Evidence: scripts/compress.py, scripts/trim_silences.py, and scripts/hw_detect.py call ffmpeg, ffprobe, and sysctl using subprocess.run().
Analysis: All subprocess calls use list-based arguments rather than shell strings. this is a secure implementation that prevents command injection vulnerabilities even when handling untrusted file paths provided as input.
[EXTERNAL_DOWNLOADS]: There are no network-based operations or remote script downloads within the skill's code or instructions.
[DATA_EXFILTRATION]: The skill does not access sensitive user files, environment variables, or credentials. It has no capability to transmit data externally.
[PROMPT_INJECTION]: The skill instructions in SKILL.md are focused on functional guidance and do not contain attempts to override agent constraints or safety protocols.

video-compressor