{"skill":{"slug":"human-approval","displayName":"Human Approval","summary":"Soft human-in-the-loop approval gate. Asks the user for confirmation before the agent executes high-risk actions like deleting files, sending emails, or runn...","tags":{"agent":"1.0.0","approval":"1.0.0","confirmation":"1.0.0","guardrails":"1.0.0","hitl":"1.0.0","human-in-the-loop":"1.0.0","latest":"1.0.0","safety":"1.0.0"},"stats":{"comments":0,"downloads":136,"installsAllTime":0,"installsCurrent":0,"stars":2,"versions":1},"createdAt":1775048267979,"updatedAt":1775051208067},"latestVersion":{"version":"1.0.0","createdAt":1775048267979,"changelog":"Initial release of the human-approval skill:\n\n- Introduces a soft human-in-the-loop approval gate for high-risk agent actions (deleting files, sending emails, destructive commands).\n- Clearly outlines categories of actions requiring explicit user confirmation, including file, communication, system, and financial operations.\n- Provides a standardized, user-friendly approval prompt format for high-risk actions.\n- Supports user responses for approve, reject, or modify actions in interactive sessions.\n- Includes commands to list, add, remove approval categories, and configure strict or off modes.\n- Describes limitations and clarifies the difference between this skill and the hard-enforcement OpenAuthority plugin.","license":"MIT-0"},"metadata":{"os":null,"systems":null},"owner":{"handle":"openauthority","userId":"s1702bpbyv1tt88m95fjpvrahn83p3z1","displayName":"OpenAuthority","image":"http://avatars.githubusercontent.com/u/272483392?s=200&v=4"},"moderation":null}