The Agent Skills Directory

[SAFE]: No malicious patterns such as data exfiltration, remote code execution, or obfuscation were detected. The skill performs localized media processing tasks and follows security best practices like quoting shell variables to prevent command injection.
[SAFE]: The skill implements human-in-the-loop checkpoints, specifically using AskUserQuestion to confirm scene detection thresholds before proceeding with frame extraction, which reduces the risk of unintended automated actions.
[SAFE]: The use of command-line tools is appropriately scoped within the allowed-tools configuration, and the data processed (video metadata and frame scores) is handled through deterministic parsing logic.

extract-frames