{"skill":{"slug":"coe-consensus","displayName":"COE Consensus","summary":"COE Consensus Engine — Cross-Model Consensus Skill for Shared World State Formation","tags":{"bft":"1.0.0","coe":"1.0.0","consensus":"1.0.0","jep":"1.0.0","latest":"1.0.0","multi-agent":"1.0.0","swarm":"1.0.0","world-model":"1.0.0"},"stats":{"comments":0,"downloads":45,"installsAllTime":0,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1777210506864,"updatedAt":1777210610576},"latestVersion":{"version":"1.0.0","createdAt":1777210506864,"changelog":"Changelog\nAll notable changes to the COE Consensus Skill will be documented in this file.\n[1.0.0] - 2026-04-26\nAdded\nInitial release of the COE Consensus Skill.\nImplemented three consensus policies from COE Protocol Section 4.2:\nSimple Majority: confirmations exceed 50% threshold.\nWeighted Trust: weighted confirmation score exceeds configurable threshold.\nByzantine Fault Tolerance (BFT): lightweight BFT variant requiring >f+1 confirmations out of >=2f+1 total.\nCore engine supporting J (Judge), V (Verify), and T (Terminate) event processing.\nConflict detection for contradictory assertions on the same subject-predicate pair.\nShared World State (SWS) generation with full provenance and consensus policy annotation.\nTermination handling: T events invalidate prior J assertions before re-consensus.\nFastAPI HTTP interface (/consensus, /health) compatible with MCP Skill calling conventions.\nFull protocol reproduction example (Appendix A robot collaboration) in examples/robot_consensus.py.\nUnit tests covering simple majority, weighted trust, BFT, termination, and unresolved scenarios.\nStandard Clawhub SKILL.md manifest with YAML frontmatter and JSON schemas.\nFeatures\nNeutral Consensus Layer: Heterogeneous world models (JEPA, Dreamer, GPT, Claude, local models) emit COE events via adapters; consensus engine produces a unified SWS.\nPolicy-Pluggable: Consensus policy configurable per session without code changes.\nAudit-Ready: All SWS records carry based_on event provenance for downstream JEP accountability tracing.\nReferences\nWang, Y. (2026). COE: Cognition-Oriented Emergence. IETF Internet-Draft.\nWang, Y. (2026). JEP: Judgment Event Protocol. IETF Internet-Draft.\nWang, Y. (2026). Target Determinability under Partial Causal Observation. Cognitive Emergence Lab.","license":"MIT-0"},"metadata":{"os":null,"systems":null},"owner":{"handle":"schchit","userId":"s1792pfv3byw2ah26z463sfgr184xckb","displayName":"JEP (Judgment Event Protocol)","image":"https://avatars.githubusercontent.com/u/10815391?v=4"},"moderation":null}