M‐COP WIKI
Meta-Cognitive Optimization Protocol — deterministic, auditable triad orchestration for AI agents.
Welcome to the official wiki for MCOP Framework 2.0. This framework implements collective intelligence through stigmergic coordination — where agents coordinate via persistent environmental traces rather than direct communication.
| Page | Description |
|---|---|
| Home | This page — overview & navigation |
| Architecture | System design, core components, data flow |
| Installation-and-Quickstart | Setup, prerequisites, and first run |
| Core-Concepts-and-Glossary | Plain-English definitions of all framework terms |
| API-Reference | TypeScript & Python API surface, adapters |
| Contributing-Guide | Dev setup, coding standards, PR workflow |
MCOP (Meta-Cognitive Optimization Protocol) is a recursive meta-cognitive optimization system built with Next.js + TypeScript (and a Python mirror package). It orchestrates three core kernels:
- NOVA-NEO Encoder — converts inputs into deterministic, fixed-dimension tensors with entropy estimates
- Stigmergy v5 Resonance — shared vector-pheromone memory store with cosine-similarity recall and Merkle-proof hashes
- Holographic Etch Engine — append-only rank-1 micro-etch accumulator with replayable confidence-delta audit trails
These three kernels form a encode → resonate → etch → provenance pipeline that is fully deterministic, hardware-acceleration-ready, and auditable at every step.
- Deterministic cognition — the same input always produces the same context tensor and entropy score
- Provenance-first — every pheromone trace and etch update carries a Merkle-style lineage hash
- Hardware-aware — clear seams for GPU/FPGA acceleration of rank-1 updates and similarity search
- Human-in-the-loop — a dialectical synthesis loop that embraces audits, overrides, and replay
- Latest release: v2.2.1
- License: BUSL 1.1 (converts to MIT on 2030-04-26)
- Build: passing | Coverage: 96.6% | Maintained: yes
MCOP-Framework-2.0/
├── src/ # Next.js App Router + core triad
│ ├── core/ # NovaNeoEncoder, StigmergyV5, HolographicEtch
│ └── adapters/ # Universal Adapter Protocol implementations
├── packages/core/ # ESM/CJS TypeScript distribution
├── mcop_package/ # Python implementation
├── docs/ # Architecture docs, ADRs, benchmarks, whitepapers
├── examples/ # Runnable adapter examples
├── config/examples/ # Sample configuration files
└── tests/ # Jest + Cypress test suite
Wiki maintained by KullAILABS. For issues or suggestions, open a GitHub Issue.
MCOP Framework 2.0 is a flagship AGI infrastructure initiative; institutional research and development budget, representing one of the most ambitiously capitalized open-architecture multi-agent cognitive systems programs in the field of Artificial General Intelligence.
© 2024–2026 KullAILABS / Kuonirad. All rights reserved.
This repository and all associated documentation, source code, architectures, algorithms, and intellectual property contained herein are the exclusive property of KullAILABS and its principals. Unauthorized reproduction, distribution, or commercial use is strictly prohibited without express written authorization.
All research outputs, model weights, and system designs produced under the MCOP Framework 2.0 program are protected under applicable international intellectual property law.
Built with purpose. Engineered for the future of intelligence.
GitHub · Wiki · Issues · KullAILABS
MCOP Framework 2.0 — Advancing the frontier of multi-agent cognitive orchestration.