The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses scripts/export_platform.py to invoke ffmpeg and ffprobe for video metadata extraction and transcoding. These operations are performed using subprocess.run and subprocess.check_output with argument lists, which prevents shell injection by ensuring user-provided paths are not interpreted as commands by a shell.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data through the video and --outdir input arguments in scripts/export_platform.py. While it lacks explicit boundary markers for these inputs, the risk is mitigated by the use of non-shell subprocess execution and the focused purpose of the scripts. Capabilities include reading file metadata and writing transcoded video files and JSON reports to the filesystem.
[SAFE]: References to external repositories in SKILL.md point to the author's own GitHub account (xianyu110), which is a legitimate vendor source for this skill. No evidence of obfuscation, credential theft, or unauthorized persistence was found during analysis.

platform-spec-export