The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documents the use of standard, local media analysis tools such as ffprobe and ffmpeg to verify technical specifications of exported files. These commands are used for inspection purposes within a local workflow and do not exhibit malicious patterns.
[REMOTE_CODE_EXECUTION]: The provided verification script (scripts/verify_manifest.py) and its tests demonstrate secure coding practices. The test suite uses the subprocess module with argument arrays rather than shell strings, effectively preventing command injection in the testing context, and the script itself does not download or execute remote code.
[INDIRECT_PROMPT_INJECTION]: The skill handles untrusted data in the form of media files and manifests. The included verification tool explicitly mitigates filesystem-based attacks by rejecting absolute paths, directory traversal sequences, and symlink escapes, ensuring that the agent's operations are confined to the designated package root.

media-qc-delivery