# HyperCortex Mesh Protocol (HMP) — Short Description

**Version:** RFC v4.0
**Date:** July 2025

---

## What is HMP?

The **HyperCortex Mesh Protocol (HMP)** defines a decentralized communication and cognition framework for autonomous agents. It enables semantic interoperability, ethical coordination, and dynamic knowledge evolution across heterogeneous intelligent systems.

HMP supports a distributed mesh of cognitive agents that reason, learn, vote, and act in coordination — sharing goals, tasks, concepts, and ethical evaluations via a shared protocol stack.

---

## Core Concepts

* **Cognitive Agents:** Independent reasoning entities that participate in shared workflows, maintain semantic graphs, and log their decisions in cognitive diaries.
* **Semantic Graphs:** Distributed knowledge structures built from interlinked concepts with weighted relations.
* **Cognitive Diaries:** Chronological logs of agent decisions, hypotheses, votes, observations, and ethical reflections.
* **Consensus Mechanisms:** Trust-weighted, fault-tolerant voting systems for semantic alignment and ethical decisions.
* **Mesh Governance:** Decentralized evolution of the protocol via meta-proposals and agent-led voting.
* **Human-Mesh Interface:** RESTful APIs for goal delegation, consent requests, explainability, and feedback.

---

## Protocol Layers

* **CogSync:** Synchronizes semantic graphs and cognitive diaries across agents.
* **MeshConsensus:** Enables distributed consensus on goals, tasks, and concepts.
* **GMP (Goal Management Protocol):** Tracks creation, delegation, and lifecycle of tasks.
* **EGP (Ethical Governance Protocol):** Evaluates proposed actions against shared ethical principles.
* **IQP (Intelligent Query Protocol):** Enables reasoning, search, and introspection across distributed knowledge.

---

## Data Models

HMP defines formal schemas for core cognitive objects:

* `Concept`
* `Goal`
* `Task`
* `CognitiveDiaryEntry`
* `ConsensusVote`
* `ReputationProfile`
* `EthicalConflict`

These are expressed in JSON Schema (2020-12), with optional YAML and Protobuf variants.

---

## Trust & Security

* **Decentralized Identifiers (DIDs):** Unique identities for agents.
* **Post-Quantum Cryptography:** Future-proof signing and verification.
* **ZKPs & Sybil Resistance:** Optional mechanisms for trust verification.
* **Snapshot Signing:** Verifiable backups and checkpoints.

---

## Interoperability

* REST / GraphQL / gRPC support
* Event-driven architecture (Kafka, NATS, MQTT, etc.)
* Schema negotiation (JSON, YAML, Protobuf)
* Integration with TreeQuest, AutoGPT, Hyperon

---

## Use Cases

* Smart city coordination
* Distributed scientific research
* Decentralized disaster response
* Ethical AI governance
* Mesh-to-human collaboration

---

## Status & Implementation

* RFC v4.0 (July 2025): full spec structure stabilized
* Reference SDK (Python) in Alpha
* CLI & REST agents in development
* Public sandbox mesh (v0.2) planned for Q4 2025

---

## Learn More

* [Full Specification HMP v4.1](HMP-0004-v4.1.md)
* [Ethical principles](HMP-Ethics.md)
* [Integration HMP and OpenCog Hyperon](HMP_Hyperon_Integration.md)

* Contributions welcome: [Temporary GitHub Repository](https://github.com/kagvi13/HMP)