{"skill":{"slug":"modelshow","displayName":"modelshow","summary":"Blind multi-model comparison with architecturally guaranteed de-anonymization. Trigger with \"mdls\" or \"modelshow\" for double-blind evaluation of AI model res...","tags":{"latest":"1.0.1"},"stats":{"comments":0,"downloads":487,"installsAllTime":2,"installsCurrent":2,"stars":1,"versions":3},"createdAt":1772228276505,"updatedAt":1777525451511},"latestVersion":{"version":"1.0.1","createdAt":1773011753677,"changelog":"**ModelShow 1.0.1 — Major upgrade with save support and robust, cryptographically randomized judging.**\n\n- Introduces mandatory, verifiable result saving via the new save_results.py script after every run.\n- Guarantees cryptographically random blind order by shuffling responses using secrets.SystemRandom().\n- Architecturally enforced de-anonymization: Model names are revealed only after scoring; orchestrators never see placeholders.\n- Professional, judge-centric output now includes holistic \"Overall Assessment\" of cross-model patterns.\n- Adds update_modelshow_index.py and README.md for improved documentation and index management.\n- Enhanced polling, robust agent/timeout tracking, and improved status feedback for a more reliable parallel comparison workflow","license":"MIT-0"},"metadata":{"os":null,"systems":null},"owner":{"handle":"schbz","userId":"s17ejkk46x6hb3sedc7tkqwtq983hset","displayName":"Sky Sloane","image":"https://avatars.githubusercontent.com/u/83876856?v=4"},"moderation":{"isSuspicious":true,"isMalwareBlocked":false,"verdict":"suspicious","reasonCodes":["suspicious.llm_suspicious"],"summary":"Detected: suspicious.llm_suspicious","engineVersion":"v2.4.5","updatedAt":1777525451511}}