{"skill":{"slug":"challenge-loop","displayName":"Challenge Loop","summary":"Challenge Loop — adversarial hardening for judgment-containing outputs. Two modes: inline self-refutation (zero latency) and independent challenger subagent....","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":72,"installsAllTime":0,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1776426378868,"updatedAt":1776426711319},"latestVersion":{"version":"1.0.0","createdAt":1776426378868,"changelog":"challenge-loop 1.0.0 — Initial Release\n\n- Introduces Challenge Loop for adversarial hardening of judgment-based outputs.\n- Supports two modes: inline self-refutation (zero latency) and subagent challenger with configurable intensity.\n- Manual trigger with natural language; user can escalate, de-escalate, or skip.\n- Subagent mode features independent challenger persona, multiple rounds, and escalating challenge intensity.\n- Canonical challenger prompt template with strict anti-recursion enforcement.\n- Platform-agnostic orchestration, with fallback to inline challenge on failure.","license":"MIT-0"},"metadata":null,"owner":{"handle":"enzowyf","userId":"s178rv61fa9jmbqwcrkzn1qncs83g805","displayName":"enzowyf","image":"https://avatars.githubusercontent.com/u/34463867?v=4"},"moderation":null}