{"skill":{"slug":"lobster-santa-method","displayName":"Lobster Santa Method","summary":"Multi-agent adversarial verification with convergence loop. Two independent review agents must both pass before output ships.","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":81,"installsAllTime":1,"installsCurrent":1,"stars":0,"versions":1},"createdAt":1775315228348,"updatedAt":1775316715286},"latestVersion":{"version":"1.0.0","createdAt":1775315228348,"changelog":"lobster-santa-method 1.0.0\n\n- Introduces a multi-agent adversarial verification framework with a phased convergence loop.\n- Implements dual independent review agents with strict context isolation before any output is shipped.\n- Outlines when to activate the skill, focusing on high-stakes, user-facing, or compliance-sensitive outputs.\n- Provides detailed architectural diagrams and step-by-step workflow for generation, independent review, verdict, and fix cycles.\n- Emphasizes the importance of objective rubric design and domain-specific extensions.\n- Describes structured review, verdict gating, iterative fixing, and escalation to human if convergence fails.","license":"MIT-0"},"metadata":null,"owner":{"handle":"wangxiaofei860208-source","userId":"s17937syhc72vf48t0a5914v8d847evs","displayName":"wangxiaofei860208-source","image":"https://avatars.githubusercontent.com/u/260913269?v=4"},"moderation":null}