{"skill":{"slug":"agent-loop","displayName":"Agent Loop","summary":"Structured Read→Plan→Execute→Verify→Report protocol for any task with side effects. Prevents false \"done\" reports, blind retries, and scope creep. Activate f...","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":318,"installsAllTime":0,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1772600337160,"updatedAt":1777525580965},"latestVersion":{"version":"1.0.0","createdAt":1772600337160,"changelog":"Initial release: Read→Plan→Execute→Verify→Report protocol for local LLMs. Enforces disciplined\n  execution loop with FORBIDDEN rules, error recovery, and scope control. Integrates with agent-task-tracker and\n  agent-step-sequencer.","license":null},"metadata":null,"owner":{"handle":"cycy2xxx","userId":"publishers:cycy2xxx","displayName":"cycy2xxx","image":"https://avatars.githubusercontent.com/u/198327825?v=4"},"moderation":null}