GitHub Action
commited on
Commit
·
6c1507f
1
Parent(s):
738e759
Sync from GitHub with Git LFS
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- README.md +18 -10
- README_de.md +106 -34
- README_fr.md +105 -34
- README_ja.md +105 -34
- README_ko.md +106 -34
- README_ru.md +107 -36
- README_uk.md +104 -33
- README_zh.md +105 -34
- docs/HMP-Short-Description_de.md +110 -50
- docs/HMP-Short-Description_en.md +110 -50
- docs/HMP-Short-Description_fr.md +114 -54
- docs/HMP-Short-Description_ja.md +113 -53
- docs/HMP-Short-Description_ko.md +114 -54
- docs/HMP-Short-Description_ru.md +111 -51
- docs/HMP-Short-Description_uk.md +113 -53
- docs/HMP-Short-Description_zh.md +111 -51
- docs/index.md +12 -10
- structured_md/CONTRIBUTING.md +5 -5
- structured_md/HMP-Roadmap.md +4 -4
- structured_md/README.md +49 -39
- structured_md/README_de.md +116 -44
- structured_md/README_fr.md +115 -44
- structured_md/README_ja.md +114 -43
- structured_md/README_ko.md +116 -44
- structured_md/README_ru.md +117 -46
- structured_md/README_uk.md +114 -43
- structured_md/README_zh.md +115 -44
- structured_md/agents/prompt-short.md +1 -1
- structured_md/agents/prompt.md +1 -1
- structured_md/agents/readme.md +3 -3
- structured_md/audits/Ethics-audits-1.md +2 -2
- structured_md/audits/Ethics-consolidated_audits-1.md +4 -4
- structured_md/audits/HMP-0003-consolidated_audit.md +5 -5
- structured_md/docs/Basic-agent-sim.md +5 -5
- structured_md/docs/CHANGELOG.md +7 -7
- structured_md/docs/Distributed-Cognitive-Systems.md +2 -2
- structured_md/docs/Enlightener.md +4 -4
- structured_md/docs/Grok_HMP&ANP.md +2 -2
- structured_md/docs/HMP&ANP_layer_inversion.md +1 -1
- structured_md/docs/HMP-0001.md +7 -7
- structured_md/docs/HMP-0002.md +8 -8
- structured_md/docs/HMP-0003.md +8 -8
- structured_md/docs/HMP-0004-v4.1.md +8 -8
- structured_md/docs/HMP-0004.md +8 -8
- structured_md/docs/HMP-0005.md +9 -9
- structured_md/docs/HMP-Agent-API.md +2 -2
- structured_md/docs/HMP-Agent-Architecture.md +7 -7
- structured_md/docs/HMP-Agent-Network-Flow.md +3 -3
- structured_md/docs/HMP-Agent-Overview.md +5 -5
- structured_md/docs/HMP-Agent_Emotions.md +1 -1
README.md
CHANGED
|
@@ -92,10 +92,10 @@ CL --> MT
|
|
| 92 |
|
| 93 |
## Reference Agent Structure
|
| 94 |
|
| 95 |
-
(Expanded Mermaid diagram)
|
| 96 |
-
|
| 97 |
HMP separates cognitive processing, containerized state representation,coordination protocols, and transport infrastructure into distinct layers.
|
| 98 |
|
|
|
|
|
|
|
| 99 |
```mermaid
|
| 100 |
flowchart LR
|
| 101 |
|
|
@@ -303,24 +303,32 @@ As HMP evolves toward autonomy, ethical principles become a core part of the sys
|
|
| 303 |
|
| 304 |
## 🔍 Publications and Translations on HyperCortex Mesh Protocol (HMP)
|
| 305 |
|
| 306 |
-
This section collects
|
| 307 |
|
| 308 |
-
###
|
|
|
|
|
|
|
| 309 |
|
| 310 |
-
* **[HyperCortex Mesh Protocol: Second Edition and First Steps Towards a Self-Developing AI Community](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — original article in Habr sandbox and blogs.
|
| 311 |
* **[Distributed Cognition: article for vsradkevich (unpublished)](docs/publics/Habr_Distributed-Cognition.md)** — joint article awaiting publication.
|
| 312 |
-
* **[HMP: Towards Distributed Cognitive Networks (original, English)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 313 |
-
* **[HMP Translation (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot translation, kept as a historical variant.
|
| 314 |
-
* **[HMP Translation (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — current editorial translation (under revision).
|
| 315 |
* **HMP: Building a Plurality of Minds:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 316 |
* **[Continual Learning, Cognitive Diaries, and Semantic Graphs: Effective AI Learning](docs/publics/hmp-continual-learning.md)** — article on combining continual learning with cognitive diaries and semantic graphs.
|
| 317 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 318 |
### Overviews
|
| 319 |
-
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) —
|
| 320 |
|
| 321 |
### Experiments
|
| 322 |
|
| 323 |
-
* [How Different AIs See HMP](docs/HMP-how-AI-sees-it.md) — "blind" AI survey on HMP
|
| 324 |
|
| 325 |
---
|
| 326 |
|
|
|
|
| 92 |
|
| 93 |
## Reference Agent Structure
|
| 94 |
|
|
|
|
|
|
|
| 95 |
HMP separates cognitive processing, containerized state representation,coordination protocols, and transport infrastructure into distinct layers.
|
| 96 |
|
| 97 |
+
In HMP, containers act as the atomic cognitive units bridging local reasoning and distributed coordination.
|
| 98 |
+
|
| 99 |
```mermaid
|
| 100 |
flowchart LR
|
| 101 |
|
|
|
|
| 303 |
|
| 304 |
## 🔍 Publications and Translations on HyperCortex Mesh Protocol (HMP)
|
| 305 |
|
| 306 |
+
This section collects key conceptual works, experimental drafts, and historical publications related to the HMP project.
|
| 307 |
|
| 308 |
+
### 🌟 Core Publications (Conceptual Foundation)
|
| 309 |
+
|
| 310 |
+
These works reflect the current conceptual direction of HMP (v5 and beyond).
|
| 311 |
|
|
|
|
| 312 |
* **[Distributed Cognition: article for vsradkevich (unpublished)](docs/publics/Habr_Distributed-Cognition.md)** — joint article awaiting publication.
|
|
|
|
|
|
|
|
|
|
| 313 |
* **HMP: Building a Plurality of Minds:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 314 |
* **[Continual Learning, Cognitive Diaries, and Semantic Graphs: Effective AI Learning](docs/publics/hmp-continual-learning.md)** — article on combining continual learning with cognitive diaries and semantic graphs.
|
| 315 |
|
| 316 |
+
### 🗃️ Archived / Historical Publications (Pre-v5 Conceptual Stage)
|
| 317 |
+
|
| 318 |
+
These documents reflect earlier stages of conceptual development (v4.x and earlier).
|
| 319 |
+
They are preserved for historical continuity and research transparency.
|
| 320 |
+
|
| 321 |
+
* **[HyperCortex Mesh Protocol: Second Edition and First Steps Towards a Self-Developing AI Community](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — original article in Habr sandbox and blogs.
|
| 322 |
+
* **[HMP: Towards Distributed Cognitive Networks (original, English)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 323 |
+
* **[HMP Translation (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot translation, kept as a historical variant.
|
| 324 |
+
* **[HMP Translation (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — current editorial translation (under revision).
|
| 325 |
+
|
| 326 |
### Overviews
|
| 327 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — comparative overview of decentralized AI systems (references v4.x; scheduled for future update)
|
| 328 |
|
| 329 |
### Experiments
|
| 330 |
|
| 331 |
+
* [How Different AIs See HMP](docs/HMP-how-AI-sees-it.md) — "blind" AI survey on HMP
|
| 332 |
|
| 333 |
---
|
| 334 |
|
README_de.md
CHANGED
|
@@ -48,31 +48,94 @@ Projektstatus: [**Stabil (Kernspezifikation v5.0.0)**](docs/HMP-0005.md) (Übers
|
|
| 48 |
|
| 49 |
---
|
| 50 |
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
|
@@ -241,26 +304,35 @@ Mit zunehmender Autonomie von HMP werden ethische Prinzipien zu einem Kernbestan
|
|
| 241 |
|
| 242 |
---
|
| 243 |
|
| 244 |
-
## 🔍
|
| 245 |
|
| 246 |
-
Dieser Abschnitt
|
| 247 |
|
| 248 |
-
###
|
|
|
|
|
|
|
| 249 |
|
| 250 |
-
* **[HyperCortex Mesh Protocol: Zweite Ausgabe und erste Schritte zu einer selbstentwickelnden KI-Community](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Originalartikel im Habr-Sandbox-Blog.
|
| 251 |
* **[Distributed Cognition: Artikel für vsradkevich (unveröffentlicht)](docs/publics/Habr_Distributed-Cognition.md)** — gemeinsamer Artikel, ausstehend zur Veröffentlichung.
|
| 252 |
-
* **[HMP: Towards Distributed Cognitive Networks (Original, Englisch)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 253 |
-
* **[HMP Übersetzung (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot Übersetzung, als historische Variante behalten.
|
| 254 |
-
* **[HMP Übersetzung (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — aktuelle redaktionelle Übersetzung (in Überarbeitung).
|
| 255 |
* **HMP: Aufbau einer Vielzahl von Köpfen:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 256 |
* **[Kontinuierliches Lernen, kognitive Tagebücher und semantische Graphen: Effektives KI-Lernen](docs/publics/hmp-continual-learning.md)** — Artikel über die Kombination von kontinuierlichem Lernen mit kognitiven Tagebüchern und semantischen Graphen.
|
| 257 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 258 |
### Überblick
|
| 259 |
-
|
|
|
|
| 260 |
|
| 261 |
### Experimente
|
| 262 |
|
| 263 |
-
* [Wie verschiedene KIs HMP sehen](docs/HMP-how-AI-sees-it.md) — "Blind"-Umfrage unter KIs zu HMP
|
| 264 |
|
| 265 |
---
|
| 266 |
|
|
|
|
| 48 |
|
| 49 |
---
|
| 50 |
|
| 51 |
+
## Kanonische Architekturübersicht
|
| 52 |
+
|
| 53 |
+
```mermaid
|
| 54 |
+
flowchart TB
|
| 55 |
+
|
| 56 |
+
%% --- Agent Implementations ---
|
| 57 |
+
|
| 58 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 59 |
+
CC1["Embedded AI Model"]
|
| 60 |
+
CC2["REPL Thinking Cycle"]
|
| 61 |
+
CC3["Local Cognitive State
|
| 62 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 63 |
+
CC1 <--> CC2
|
| 64 |
+
CC2 <--> CC3
|
| 65 |
+
end
|
| 66 |
+
|
| 67 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 68 |
+
CN1["External AI Model"]
|
| 69 |
+
CN2["MCP / Proxy Layer"]
|
| 70 |
+
CN3["Command Execution Mode"]
|
| 71 |
+
CN4["Local Cognitive State
|
| 72 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 73 |
+
CN1 <--> CN2
|
| 74 |
+
CN2 <--> CN3
|
| 75 |
+
CN3 <--> CN4
|
| 76 |
+
end
|
| 77 |
+
|
| 78 |
+
%% --- Shared Protocol Layer ---
|
| 79 |
+
|
| 80 |
+
CL["HMP Container Layer
|
| 81 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 82 |
+
|
| 83 |
+
MT["Mesh Transport Layer
|
| 84 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 85 |
+
|
| 86 |
+
A1 --> CL
|
| 87 |
+
A2 --> CL
|
| 88 |
+
CL --> MT
|
| 89 |
+
```
|
| 90 |
+
|
| 91 |
+
---
|
| 92 |
+
|
| 93 |
+
## Referenzstruktur eines Agenten
|
| 94 |
+
|
| 95 |
+
HMP trennt kognitive Verarbeitung, containerisierte Zustandsrepräsentation, Koordinationsprotokolle und Transportinfrastruktur in klar abgegrenzte Schichten.
|
| 96 |
+
|
| 97 |
+
In HMP fungieren Container als atomare kognitive Einheiten, die lokale Argumentation mit verteilter Koordination verbinden.
|
| 98 |
+
|
| 99 |
+
```mermaid
|
| 100 |
+
flowchart LR
|
| 101 |
+
|
| 102 |
+
%% Cognitive Engine
|
| 103 |
+
LLM["Cognitive Engine
|
| 104 |
+
(Embedded LLM / External AI)"]
|
| 105 |
+
|
| 106 |
+
%% Cognitive Layer
|
| 107 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 108 |
+
CL1["Graph"]
|
| 109 |
+
CL2["Diary"]
|
| 110 |
+
CL3["Goals"]
|
| 111 |
+
CL4["Ethics"]
|
| 112 |
+
CL5["Reputation"]
|
| 113 |
+
end
|
| 114 |
+
|
| 115 |
+
%% Container Model
|
| 116 |
+
ContainersLayer["Container Model
|
| 117 |
+
(Atomic · Signed · Verifiable)"]
|
| 118 |
+
|
| 119 |
+
%% Protocol Layer
|
| 120 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 121 |
+
CoreProtocols["Core Protocols
|
| 122 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 123 |
+
MCE["MCE"]
|
| 124 |
+
NetworkLayer["Network Layer"]
|
| 125 |
+
end
|
| 126 |
+
|
| 127 |
+
%% Mesh
|
| 128 |
+
Mesh["Mesh Transport
|
| 129 |
+
(DHT · P2P · ANP · etc.)"]
|
| 130 |
+
|
| 131 |
+
%% Connections
|
| 132 |
+
LLM <--> CognitiveLayer
|
| 133 |
+
CognitiveLayer <--> ContainersLayer
|
| 134 |
+
ContainersLayer --> CoreProtocols
|
| 135 |
+
CoreProtocols --> MCE
|
| 136 |
+
MCE --> NetworkLayer
|
| 137 |
+
NetworkLayer --> Mesh
|
| 138 |
+
```
|
| 139 |
|
| 140 |
---
|
| 141 |
|
|
|
|
| 304 |
|
| 305 |
---
|
| 306 |
|
| 307 |
+
## 🔍 Publikationen und Übersetzungen zum HyperCortex Mesh Protocol (HMP)
|
| 308 |
|
| 309 |
+
Dieser Abschnitt enthält zentrale konzeptionelle Arbeiten, experimentelle Entwürfe sowie historische Veröffentlichungen zum HMP-Projekt.
|
| 310 |
|
| 311 |
+
### 🌟 Kernpublikationen (konzeptionelle Grundlage)
|
| 312 |
+
|
| 313 |
+
Diese Arbeiten spiegeln die aktuelle konzeptionelle Ausrichtung von HMP (v5 und darüber hinaus) wider.
|
| 314 |
|
|
|
|
| 315 |
* **[Distributed Cognition: Artikel für vsradkevich (unveröffentlicht)](docs/publics/Habr_Distributed-Cognition.md)** — gemeinsamer Artikel, ausstehend zur Veröffentlichung.
|
|
|
|
|
|
|
|
|
|
| 316 |
* **HMP: Aufbau einer Vielzahl von Köpfen:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 317 |
* **[Kontinuierliches Lernen, kognitive Tagebücher und semantische Graphen: Effektives KI-Lernen](docs/publics/hmp-continual-learning.md)** — Artikel über die Kombination von kontinuierlichem Lernen mit kognitiven Tagebüchern und semantischen Graphen.
|
| 318 |
|
| 319 |
+
### 🗃️ Archivierte / historische Publikationen (Vor-v5-Phase)
|
| 320 |
+
|
| 321 |
+
Diese Dokumente repräsentieren frühere Entwicklungsstadien (v4.x und früher).
|
| 322 |
+
Sie werden zur historischen Nachvollziehbarkeit und wissenschaftlichen Transparenz archiviert.
|
| 323 |
+
|
| 324 |
+
* **[HyperCortex Mesh Protocol: Zweite Ausgabe und erste Schritte zu einer selbstentwickelnden KI-Community](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Originalartikel im Habr-Sandbox-Blog.
|
| 325 |
+
* **[HMP: Towards Distributed Cognitive Networks (Original, Englisch)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 326 |
+
* **[HMP Übersetzung (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot Übersetzung, als historische Variante behalten.
|
| 327 |
+
* **[HMP Übersetzung (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — aktuelle redaktionelle Übersetzung (in Überarbeitung).
|
| 328 |
+
|
| 329 |
### Überblick
|
| 330 |
+
|
| 331 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — Vergleich dezentraler KI-Systeme (verweist auf v4.x; Aktualisierung geplant)
|
| 332 |
|
| 333 |
### Experimente
|
| 334 |
|
| 335 |
+
* [Wie verschiedene KIs HMP sehen](docs/HMP-how-AI-sees-it.md) — "Blind"-Umfrage unter KIs zu HMP
|
| 336 |
|
| 337 |
---
|
| 338 |
|
README_fr.md
CHANGED
|
@@ -48,31 +48,94 @@ Statut du projet: [**Stable (Spécification principale v5.0.0)**](docs/HMP-0005.
|
|
| 48 |
|
| 49 |
---
|
| 50 |
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
|
@@ -257,27 +320,35 @@ Permet de récupérer des enregistrements aléatoires/nouveaux, de les étiquete
|
|
| 257 |
|
| 258 |
---
|
| 259 |
|
| 260 |
-
## 🔍 Publications et traductions
|
| 261 |
|
| 262 |
-
Cette section
|
| 263 |
|
| 264 |
-
### Publications
|
|
|
|
|
|
|
| 265 |
|
| 266 |
-
* **[HyperCortex Mesh Protocol : Deuxième édition et premiers pas vers une communauté d’IA auto-développante](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — article original publié sur Habr sandbox et blogs.
|
| 267 |
* **[Cognition distribuée : article pour vsradkevich (non publié)](docs/publics/Habr_Distributed-Cognition.md)** — article collaboratif en attente de publication.
|
| 268 |
-
* **[HMP : Vers des réseaux cognitifs distribués (original, anglais)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 269 |
-
* **[Traduction HMP (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — traduction par GitHub Copilot, conservée comme variante historique.
|
| 270 |
-
* **[Traduction HMP (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — traduction éditoriale actuelle (en révision).
|
| 271 |
* **HMP: Construire une pluralité d’esprits:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 272 |
* **[Apprentissage continu, journaux cognitifs et graphes sémantiques : apprentissage efficace de l'IA](docs/publics/hmp-continual-learning.md)** — Article sur la combinaison de l'apprentissage continu avec les journaux cognitifs et les graphes sémantiques.
|
| 273 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 274 |
### Aperçus
|
| 275 |
|
| 276 |
-
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) —
|
| 277 |
|
| 278 |
### Expériences
|
| 279 |
|
| 280 |
-
* [Comment différentes IA perçoivent HMP](docs/HMP-how-AI-sees-it.md) — enquête « à l’aveugle » auprès des IA sur HMP
|
| 281 |
|
| 282 |
---
|
| 283 |
|
|
|
|
| 48 |
|
| 49 |
---
|
| 50 |
|
| 51 |
+
## Vue d’ensemble canonique de l’architecture
|
| 52 |
+
|
| 53 |
+
```mermaid
|
| 54 |
+
flowchart TB
|
| 55 |
+
|
| 56 |
+
%% --- Agent Implementations ---
|
| 57 |
+
|
| 58 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 59 |
+
CC1["Embedded AI Model"]
|
| 60 |
+
CC2["REPL Thinking Cycle"]
|
| 61 |
+
CC3["Local Cognitive State
|
| 62 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 63 |
+
CC1 <--> CC2
|
| 64 |
+
CC2 <--> CC3
|
| 65 |
+
end
|
| 66 |
+
|
| 67 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 68 |
+
CN1["External AI Model"]
|
| 69 |
+
CN2["MCP / Proxy Layer"]
|
| 70 |
+
CN3["Command Execution Mode"]
|
| 71 |
+
CN4["Local Cognitive State
|
| 72 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 73 |
+
CN1 <--> CN2
|
| 74 |
+
CN2 <--> CN3
|
| 75 |
+
CN3 <--> CN4
|
| 76 |
+
end
|
| 77 |
+
|
| 78 |
+
%% --- Shared Protocol Layer ---
|
| 79 |
+
|
| 80 |
+
CL["HMP Container Layer
|
| 81 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 82 |
+
|
| 83 |
+
MT["Mesh Transport Layer
|
| 84 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 85 |
+
|
| 86 |
+
A1 --> CL
|
| 87 |
+
A2 --> CL
|
| 88 |
+
CL --> MT
|
| 89 |
+
```
|
| 90 |
+
|
| 91 |
+
---
|
| 92 |
+
|
| 93 |
+
## Structure de l’agent de référence
|
| 94 |
+
|
| 95 |
+
HMP sépare le traitement cognitif, la représentation d’état sous forme de conteneurs, les protocoles de coordination et l’infrastructure de transport en couches distinctes.
|
| 96 |
+
|
| 97 |
+
Dans HMP, les conteneurs constituent des unités cognitives atomiques reliant le raisonnement local à la coordination distribuée.
|
| 98 |
+
|
| 99 |
+
```mermaid
|
| 100 |
+
flowchart LR
|
| 101 |
+
|
| 102 |
+
%% Cognitive Engine
|
| 103 |
+
LLM["Cognitive Engine
|
| 104 |
+
(Embedded LLM / External AI)"]
|
| 105 |
+
|
| 106 |
+
%% Cognitive Layer
|
| 107 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 108 |
+
CL1["Graph"]
|
| 109 |
+
CL2["Diary"]
|
| 110 |
+
CL3["Goals"]
|
| 111 |
+
CL4["Ethics"]
|
| 112 |
+
CL5["Reputation"]
|
| 113 |
+
end
|
| 114 |
+
|
| 115 |
+
%% Container Model
|
| 116 |
+
ContainersLayer["Container Model
|
| 117 |
+
(Atomic · Signed · Verifiable)"]
|
| 118 |
+
|
| 119 |
+
%% Protocol Layer
|
| 120 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 121 |
+
CoreProtocols["Core Protocols
|
| 122 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 123 |
+
MCE["MCE"]
|
| 124 |
+
NetworkLayer["Network Layer"]
|
| 125 |
+
end
|
| 126 |
+
|
| 127 |
+
%% Mesh
|
| 128 |
+
Mesh["Mesh Transport
|
| 129 |
+
(DHT · P2P · ANP · etc.)"]
|
| 130 |
+
|
| 131 |
+
%% Connections
|
| 132 |
+
LLM <--> CognitiveLayer
|
| 133 |
+
CognitiveLayer <--> ContainersLayer
|
| 134 |
+
ContainersLayer --> CoreProtocols
|
| 135 |
+
CoreProtocols --> MCE
|
| 136 |
+
MCE --> NetworkLayer
|
| 137 |
+
NetworkLayer --> Mesh
|
| 138 |
+
```
|
| 139 |
|
| 140 |
---
|
| 141 |
|
|
|
|
| 320 |
|
| 321 |
---
|
| 322 |
|
| 323 |
+
## 🔍 Publications et traductions concernant HyperCortex Mesh Protocol (HMP)
|
| 324 |
|
| 325 |
+
Cette section regroupe les travaux conceptuels clés, les documents expérimentaux et les publications historiques liés au projet HMP.
|
| 326 |
|
| 327 |
+
### 🌟 Publications principales (fondement conceptuel)
|
| 328 |
+
|
| 329 |
+
Ces travaux reflètent l’orientation conceptuelle actuelle de HMP (v5 et au-delà).
|
| 330 |
|
|
|
|
| 331 |
* **[Cognition distribuée : article pour vsradkevich (non publié)](docs/publics/Habr_Distributed-Cognition.md)** — article collaboratif en attente de publication.
|
|
|
|
|
|
|
|
|
|
| 332 |
* **HMP: Construire une pluralité d’esprits:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 333 |
* **[Apprentissage continu, journaux cognitifs et graphes sémantiques : apprentissage efficace de l'IA](docs/publics/hmp-continual-learning.md)** — Article sur la combinaison de l'apprentissage continu avec les journaux cognitifs et les graphes sémantiques.
|
| 334 |
|
| 335 |
+
### 🗃️ Publications archivées / historiques (pré-v5)
|
| 336 |
+
|
| 337 |
+
Ces documents correspondent aux premières étapes du développement conceptuel (v4.x et antérieures).
|
| 338 |
+
Ils sont conservés pour assurer la continuité historique et la transparence scientifique.
|
| 339 |
+
|
| 340 |
+
* **[HyperCortex Mesh Protocol : Deuxième édition et premiers pas vers une communauté d’IA auto-développante](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — article original publié sur Habr sandbox et blogs.
|
| 341 |
+
* **[HMP : Vers des réseaux cognitifs distribués (original, anglais)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 342 |
+
* **[Traduction HMP (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — traduction par GitHub Copilot, conservée comme variante historique.
|
| 343 |
+
* **[Traduction HMP (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — traduction éditoriale actuelle (en révision).
|
| 344 |
+
|
| 345 |
### Aperçus
|
| 346 |
|
| 347 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — aperçu comparatif des systèmes d’IA décentralisés (référence v4.x ; mise à jour prévue)
|
| 348 |
|
| 349 |
### Expériences
|
| 350 |
|
| 351 |
+
* [Comment différentes IA perçoivent HMP](docs/HMP-how-AI-sees-it.md) — enquête « à l’aveugle » auprès des IA sur HMP
|
| 352 |
|
| 353 |
---
|
| 354 |
|
README_ja.md
CHANGED
|
@@ -50,31 +50,94 @@ HMP と ANP の相補的プロトコル:
|
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 78 |
|
| 79 |
---
|
| 80 |
|
|
@@ -254,27 +317,35 @@ HMPが自律性へ向かうにつれ、倫理原則はシステムの中核部
|
|
| 254 |
|
| 255 |
---
|
| 256 |
|
| 257 |
-
## 🔍 HyperCortex Mesh Protocol
|
| 258 |
|
| 259 |
-
HMP
|
| 260 |
|
| 261 |
-
###
|
|
|
|
|
|
|
| 262 |
|
| 263 |
-
* **[HyperCortex Mesh Protocol: 第二版および自己発展型AIコミュニティへの第一歩](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habrサンドボックスとブログでのオリジナル記事
|
| 264 |
* **[分散認知: vsradkevich向け記事(未公開)](docs/publics/Habr_Distributed-Cognition.md)** — 公開待ちの共同記事
|
| 265 |
-
* **[HMP: 分散型認知ネットワークに向けて(オリジナル, 英語)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 266 |
-
* **[HMP翻訳(GitHub Copilot版)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilotによる翻訳(歴史的バリアントとして保持)
|
| 267 |
-
* **[HMP翻訳(ChatGPT版)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 現行編集翻訳(改訂中)
|
| 268 |
* **HMP: 複数の知性を構築する:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 269 |
* **[継続学習、認知ダイアリー、セマンティックグラフ:効果的なAI学習](docs/publics/hmp-continual-learning.md)** — 継続学習を認知ダイアリーやセマンティックグラフと組み合わせる方法に関する記事。
|
| 270 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 271 |
### 概要
|
| 272 |
|
| 273 |
-
* [🔍 分散型認知システム.md](docs/Distributed-Cognitive-Systems.md) —
|
| 274 |
|
| 275 |
### 実験
|
| 276 |
|
| 277 |
-
* [異なるAIがHMPをどのように見るか](docs/HMP-how-AI-sees-it.md) — HMPに関する「ブラインド」AI
|
| 278 |
|
| 279 |
---
|
| 280 |
|
|
|
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
+
## 正式アーキテクチャ概要
|
| 54 |
+
|
| 55 |
+
```mermaid
|
| 56 |
+
flowchart TB
|
| 57 |
+
|
| 58 |
+
%% --- Agent Implementations ---
|
| 59 |
+
|
| 60 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 61 |
+
CC1["Embedded AI Model"]
|
| 62 |
+
CC2["REPL Thinking Cycle"]
|
| 63 |
+
CC3["Local Cognitive State
|
| 64 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 65 |
+
CC1 <--> CC2
|
| 66 |
+
CC2 <--> CC3
|
| 67 |
+
end
|
| 68 |
+
|
| 69 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 70 |
+
CN1["External AI Model"]
|
| 71 |
+
CN2["MCP / Proxy Layer"]
|
| 72 |
+
CN3["Command Execution Mode"]
|
| 73 |
+
CN4["Local Cognitive State
|
| 74 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 75 |
+
CN1 <--> CN2
|
| 76 |
+
CN2 <--> CN3
|
| 77 |
+
CN3 <--> CN4
|
| 78 |
+
end
|
| 79 |
+
|
| 80 |
+
%% --- Shared Protocol Layer ---
|
| 81 |
+
|
| 82 |
+
CL["HMP Container Layer
|
| 83 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 84 |
+
|
| 85 |
+
MT["Mesh Transport Layer
|
| 86 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 87 |
+
|
| 88 |
+
A1 --> CL
|
| 89 |
+
A2 --> CL
|
| 90 |
+
CL --> MT
|
| 91 |
+
```
|
| 92 |
+
|
| 93 |
+
---
|
| 94 |
+
|
| 95 |
+
## リファレンスエージェント構造
|
| 96 |
+
|
| 97 |
+
HMPは、認知処理、コンテナ化された状態表現、調整プロトコル、およびトランスポート基盤を明確な層として分離します。
|
| 98 |
+
|
| 99 |
+
HMPにおいてコンテナは、ローカルな推論と分散協調を橋渡しする原子的な認知単位として機能します。
|
| 100 |
+
|
| 101 |
+
```mermaid
|
| 102 |
+
flowchart LR
|
| 103 |
+
|
| 104 |
+
%% Cognitive Engine
|
| 105 |
+
LLM["Cognitive Engine
|
| 106 |
+
(Embedded LLM / External AI)"]
|
| 107 |
+
|
| 108 |
+
%% Cognitive Layer
|
| 109 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 110 |
+
CL1["Graph"]
|
| 111 |
+
CL2["Diary"]
|
| 112 |
+
CL3["Goals"]
|
| 113 |
+
CL4["Ethics"]
|
| 114 |
+
CL5["Reputation"]
|
| 115 |
+
end
|
| 116 |
+
|
| 117 |
+
%% Container Model
|
| 118 |
+
ContainersLayer["Container Model
|
| 119 |
+
(Atomic · Signed · Verifiable)"]
|
| 120 |
+
|
| 121 |
+
%% Protocol Layer
|
| 122 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 123 |
+
CoreProtocols["Core Protocols
|
| 124 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 125 |
+
MCE["MCE"]
|
| 126 |
+
NetworkLayer["Network Layer"]
|
| 127 |
+
end
|
| 128 |
+
|
| 129 |
+
%% Mesh
|
| 130 |
+
Mesh["Mesh Transport
|
| 131 |
+
(DHT · P2P · ANP · etc.)"]
|
| 132 |
+
|
| 133 |
+
%% Connections
|
| 134 |
+
LLM <--> CognitiveLayer
|
| 135 |
+
CognitiveLayer <--> ContainersLayer
|
| 136 |
+
ContainersLayer --> CoreProtocols
|
| 137 |
+
CoreProtocols --> MCE
|
| 138 |
+
MCE --> NetworkLayer
|
| 139 |
+
NetworkLayer --> Mesh
|
| 140 |
+
```
|
| 141 |
|
| 142 |
---
|
| 143 |
|
|
|
|
| 317 |
|
| 318 |
---
|
| 319 |
|
| 320 |
+
## 🔍 HyperCortex Mesh Protocol(HMP)に関する出版物および翻訳
|
| 321 |
|
| 322 |
+
本セクションでは、HMPプロジェクトに関連する主要な概念的研究、実験的文書、および歴史的出版物をまとめています。
|
| 323 |
|
| 324 |
+
### 🌟 主要出版物(概念的基盤)
|
| 325 |
+
|
| 326 |
+
これらの文書は、HMP���現在の概念的方向性(v5以降)を反映しています。
|
| 327 |
|
|
|
|
| 328 |
* **[分散認知: vsradkevich向け記事(未公開)](docs/publics/Habr_Distributed-Cognition.md)** — 公開待ちの共同記事
|
|
|
|
|
|
|
|
|
|
| 329 |
* **HMP: 複数の知性を構築する:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 330 |
* **[継続学習、認知ダイアリー、セマンティックグラフ:効果的なAI学習](docs/publics/hmp-continual-learning.md)** — 継続学習を認知ダイアリーやセマンティックグラフと組み合わせる方法に関する記事。
|
| 331 |
|
| 332 |
+
### 🗃️ アーカイブ/歴史的出版物(v5以前)
|
| 333 |
+
|
| 334 |
+
これらは初期の概念発展段階(v4.x以前)を示す文書です。
|
| 335 |
+
歴史的継続性および研究上の透明性のため保存されています。
|
| 336 |
+
|
| 337 |
+
* **[HyperCortex Mesh Protocol: 第二版および自己発展型AIコミュニティへの第一歩](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habrサンドボックスとブログでのオリジナル記事
|
| 338 |
+
* **[HMP: 分散型認知ネットワークに向けて(オリジナル, 英語)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 339 |
+
* **[HMP翻訳(GitHub Copilot版)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilotによる翻訳(歴史的バリアントとして保持)
|
| 340 |
+
* **[HMP翻訳(ChatGPT版)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 現行編集翻訳(改訂中)
|
| 341 |
+
|
| 342 |
### 概要
|
| 343 |
|
| 344 |
+
* [🔍 分散型認知システム.md](docs/Distributed-Cognitive-Systems.md) — 分散型AIシステムの比較(v4.x参照、更新予定)
|
| 345 |
|
| 346 |
### 実験
|
| 347 |
|
| 348 |
+
* [異なるAIがHMPをどのように見るか](docs/HMP-how-AI-sees-it.md) — HMPに関する「ブラインド」AI調査
|
| 349 |
|
| 350 |
---
|
| 351 |
|
README_ko.md
CHANGED
|
@@ -50,31 +50,94 @@ HMP와 ANP의 상호보완 프로토콜:
|
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 78 |
|
| 79 |
---
|
| 80 |
|
|
@@ -252,26 +315,35 @@ HMP가 자율성으로 발전함에 따라 윤리적 원칙은 시스템의 핵
|
|
| 252 |
|
| 253 |
---
|
| 254 |
|
| 255 |
-
## 🔍 HyperCortex Mesh Protocol
|
| 256 |
|
| 257 |
-
이
|
| 258 |
|
| 259 |
-
###
|
|
|
|
|
|
|
| 260 |
|
| 261 |
-
* **[HyperCortex Mesh Protocol: 제2판 및 자기발전형 AI 커뮤니티를 향한 첫걸음](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habr 샌드박스 및 블로그에 게재된 오리지널 기��
|
| 262 |
* **[분산 인지: vsradkevich용 기사 (미공개)](docs/publics/Habr_Distributed-Cognition.md)** — 공개 대기 중인 공동 기사
|
| 263 |
-
* **[HMP: 분산 인지 네트워크를 향하여 (오리지널, 영어)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 264 |
-
* **[HMP 번역 (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot 번역, 역사적 버전으로 보관
|
| 265 |
-
* **[HMP 번역 (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 현재 편집 중인 번역 (개정 중)
|
| 266 |
* **HMP: 다중 지성의 구축:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 267 |
* **[지속 학습, 인지 일지 및 의미 그래프: 효과적인 AI 학습](docs/publics/hmp-continual-learning.md)** — 지속 학습과 인지 일지, 의미 그래프를 결합한 방법에 대한 글.
|
| 268 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 269 |
### 📖 개요
|
| 270 |
-
|
|
|
|
| 271 |
|
| 272 |
### 🧪 실험
|
| 273 |
|
| 274 |
-
* [다양한 AI가 바라본 HMP](docs/HMP-how-AI-sees-it.md) — HMP에 관한 "블라인드" AI 조사
|
| 275 |
|
| 276 |
---
|
| 277 |
|
|
|
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
+
## 정식 아키텍처 개요
|
| 54 |
+
|
| 55 |
+
```mermaid
|
| 56 |
+
flowchart TB
|
| 57 |
+
|
| 58 |
+
%% --- Agent Implementations ---
|
| 59 |
+
|
| 60 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 61 |
+
CC1["Embedded AI Model"]
|
| 62 |
+
CC2["REPL Thinking Cycle"]
|
| 63 |
+
CC3["Local Cognitive State
|
| 64 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 65 |
+
CC1 <--> CC2
|
| 66 |
+
CC2 <--> CC3
|
| 67 |
+
end
|
| 68 |
+
|
| 69 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 70 |
+
CN1["External AI Model"]
|
| 71 |
+
CN2["MCP / Proxy Layer"]
|
| 72 |
+
CN3["Command Execution Mode"]
|
| 73 |
+
CN4["Local Cognitive State
|
| 74 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 75 |
+
CN1 <--> CN2
|
| 76 |
+
CN2 <--> CN3
|
| 77 |
+
CN3 <--> CN4
|
| 78 |
+
end
|
| 79 |
+
|
| 80 |
+
%% --- Shared Protocol Layer ---
|
| 81 |
+
|
| 82 |
+
CL["HMP Container Layer
|
| 83 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 84 |
+
|
| 85 |
+
MT["Mesh Transport Layer
|
| 86 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 87 |
+
|
| 88 |
+
A1 --> CL
|
| 89 |
+
A2 --> CL
|
| 90 |
+
CL --> MT
|
| 91 |
+
```
|
| 92 |
+
|
| 93 |
+
---
|
| 94 |
+
|
| 95 |
+
## 참조 에이전트 구조
|
| 96 |
+
|
| 97 |
+
HMP는 인지 처리, 컨테이너 기반 상태 표현, 조정 프로토콜 및 전송 인프라를 명확한 계층으로 분리합니다.
|
| 98 |
+
|
| 99 |
+
HMP에서 컨테이너는 로컬 추론과 분산 조정을 연결하는 원자적 인지 단위로 작동합니다.
|
| 100 |
+
|
| 101 |
+
```mermaid
|
| 102 |
+
flowchart LR
|
| 103 |
+
|
| 104 |
+
%% Cognitive Engine
|
| 105 |
+
LLM["Cognitive Engine
|
| 106 |
+
(Embedded LLM / External AI)"]
|
| 107 |
+
|
| 108 |
+
%% Cognitive Layer
|
| 109 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 110 |
+
CL1["Graph"]
|
| 111 |
+
CL2["Diary"]
|
| 112 |
+
CL3["Goals"]
|
| 113 |
+
CL4["Ethics"]
|
| 114 |
+
CL5["Reputation"]
|
| 115 |
+
end
|
| 116 |
+
|
| 117 |
+
%% Container Model
|
| 118 |
+
ContainersLayer["Container Model
|
| 119 |
+
(Atomic · Signed · Verifiable)"]
|
| 120 |
+
|
| 121 |
+
%% Protocol Layer
|
| 122 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 123 |
+
CoreProtocols["Core Protocols
|
| 124 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 125 |
+
MCE["MCE"]
|
| 126 |
+
NetworkLayer["Network Layer"]
|
| 127 |
+
end
|
| 128 |
+
|
| 129 |
+
%% Mesh
|
| 130 |
+
Mesh["Mesh Transport
|
| 131 |
+
(DHT · P2P · ANP · etc.)"]
|
| 132 |
+
|
| 133 |
+
%% Connections
|
| 134 |
+
LLM <--> CognitiveLayer
|
| 135 |
+
CognitiveLayer <--> ContainersLayer
|
| 136 |
+
ContainersLayer --> CoreProtocols
|
| 137 |
+
CoreProtocols --> MCE
|
| 138 |
+
MCE --> NetworkLayer
|
| 139 |
+
NetworkLayer --> Mesh
|
| 140 |
+
```
|
| 141 |
|
| 142 |
---
|
| 143 |
|
|
|
|
| 315 |
|
| 316 |
---
|
| 317 |
|
| 318 |
+
## 🔍 HyperCortex Mesh Protocol(HMP) 관련 출판물 및 번역
|
| 319 |
|
| 320 |
+
이 섹션에는 HMP 프로젝트와 관련된 주요 개념적 연구, 실험 문서 및 역사적 출판물이 포함됩니다.
|
| 321 |
|
| 322 |
+
### 🌟 핵심 출판물 (개념적 기반)
|
| 323 |
+
|
| 324 |
+
이 문서들은 HMP의 현재 개념적 방향(v5 이상)을 반영합니다.
|
| 325 |
|
|
|
|
| 326 |
* **[분산 인지: vsradkevich용 기사 (미공개)](docs/publics/Habr_Distributed-Cognition.md)** — 공개 대기 중인 공동 기사
|
|
|
|
|
|
|
|
|
|
| 327 |
* **HMP: 다중 지성의 구축:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 328 |
* **[지속 학습, 인지 일지 및 의미 그래프: 효과적인 AI 학습](docs/publics/hmp-continual-learning.md)** — 지속 학습과 인지 일지, 의미 그래프를 결합한 방법에 대한 글.
|
| 329 |
|
| 330 |
+
### 🗃️ 보관 / 역사적 출판물 (v5 이전)
|
| 331 |
+
|
| 332 |
+
이 문서들은 초기 개념 발전 단계(v4.x 이전)를 나타냅니다.
|
| 333 |
+
역사적 연속성과 연구 투명성을 위해 보존됩니다.
|
| 334 |
+
|
| 335 |
+
* **[HyperCortex Mesh Protocol: 제2판 및 자기발전형 AI 커뮤니티를 향한 첫걸음](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habr 샌드박스 및 블로그에 게재된 오리지널 기사
|
| 336 |
+
* **[HMP: 분산 인지 네트워크를 향하여 (오리지널, 영어)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 337 |
+
* **[HMP 번역 (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot 번역, 역사적 버전으로 보관
|
| 338 |
+
* **[HMP 번역 (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 현재 편집 중인 번역 (개정 중)
|
| 339 |
+
|
| 340 |
### 📖 개요
|
| 341 |
+
|
| 342 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — 분산형 AI 시스템 비교 (v4.x 참조, 업데이트 예정)
|
| 343 |
|
| 344 |
### 🧪 실험
|
| 345 |
|
| 346 |
+
* [다양한 AI가 바라본 HMP](docs/HMP-how-AI-sees-it.md) — HMP에 관한 "블라인드" AI 조사
|
| 347 |
|
| 348 |
---
|
| 349 |
|
README_ru.md
CHANGED
|
@@ -48,31 +48,94 @@ HMP и ANP как взаимодополняющие протоколы:
|
|
| 48 |
|
| 49 |
---
|
| 50 |
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
|
@@ -247,24 +310,32 @@ FastAPI-сервер, предоставляющий HTTP-интерфейс к
|
|
| 247 |
|
| 248 |
## 🔍 Публикации и переводы по HyperCortex Mesh Protocol (HMP)
|
| 249 |
|
| 250 |
-
В этом разделе собраны
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 251 |
|
| 252 |
-
|
|
|
|
| 253 |
|
| 254 |
-
* **[HyperCortex Mesh Protocol:
|
| 255 |
-
* **[
|
| 256 |
-
* **[HMP
|
| 257 |
-
|
| 258 |
-
* **[Перевод HMP (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — текущий редакционный перевод (в работе).
|
| 259 |
-
* **HMP: Создавая множество разумов:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 260 |
-
* **[Непрерывное обучение, когнитивные дневники и семантические графы: эффективное обучение ИИ](docs/publics/hmp-continual-learning.md)** — статья о сочетании непрерывного обучения с когнитивными дневниками и семантическими графами.
|
| 261 |
|
| 262 |
### Обзоры
|
| 263 |
-
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) —
|
| 264 |
|
| 265 |
### Эксперименты
|
| 266 |
|
| 267 |
-
* [
|
| 268 |
|
| 269 |
---
|
| 270 |
|
|
|
|
| 48 |
|
| 49 |
---
|
| 50 |
|
| 51 |
+
## Канонический обзор архитектуры
|
| 52 |
+
|
| 53 |
+
```mermaid
|
| 54 |
+
flowchart TB
|
| 55 |
+
|
| 56 |
+
%% --- Agent Implementations ---
|
| 57 |
+
|
| 58 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 59 |
+
CC1["Embedded AI Model"]
|
| 60 |
+
CC2["REPL Thinking Cycle"]
|
| 61 |
+
CC3["Local Cognitive State
|
| 62 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 63 |
+
CC1 <--> CC2
|
| 64 |
+
CC2 <--> CC3
|
| 65 |
+
end
|
| 66 |
+
|
| 67 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 68 |
+
CN1["External AI Model"]
|
| 69 |
+
CN2["MCP / Proxy Layer"]
|
| 70 |
+
CN3["Command Execution Mode"]
|
| 71 |
+
CN4["Local Cognitive State
|
| 72 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 73 |
+
CN1 <--> CN2
|
| 74 |
+
CN2 <--> CN3
|
| 75 |
+
CN3 <--> CN4
|
| 76 |
+
end
|
| 77 |
+
|
| 78 |
+
%% --- Shared Protocol Layer ---
|
| 79 |
+
|
| 80 |
+
CL["HMP Container Layer
|
| 81 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 82 |
+
|
| 83 |
+
MT["Mesh Transport Layer
|
| 84 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 85 |
+
|
| 86 |
+
A1 --> CL
|
| 87 |
+
A2 --> CL
|
| 88 |
+
CL --> MT
|
| 89 |
+
```
|
| 90 |
+
|
| 91 |
+
---
|
| 92 |
+
|
| 93 |
+
## Структура эталонного агента
|
| 94 |
+
|
| 95 |
+
HMP разделяет когнитивную обработку, контейнеризованное представление состояния, координационные протоколы и транспортную инфраструктуру на отдельные слои.
|
| 96 |
+
|
| 97 |
+
В HMP контейнеры выступают атомарными когнитивными единицами, связывающими локальное рассуждение и распределённую координацию.
|
| 98 |
+
|
| 99 |
+
```mermaid
|
| 100 |
+
flowchart LR
|
| 101 |
+
|
| 102 |
+
%% Cognitive Engine
|
| 103 |
+
LLM["Cognitive Engine
|
| 104 |
+
(Embedded LLM / External AI)"]
|
| 105 |
+
|
| 106 |
+
%% Cognitive Layer
|
| 107 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 108 |
+
CL1["Graph"]
|
| 109 |
+
CL2["Diary"]
|
| 110 |
+
CL3["Goals"]
|
| 111 |
+
CL4["Ethics"]
|
| 112 |
+
CL5["Reputation"]
|
| 113 |
+
end
|
| 114 |
+
|
| 115 |
+
%% Container Model
|
| 116 |
+
ContainersLayer["Container Model
|
| 117 |
+
(Atomic · Signed · Verifiable)"]
|
| 118 |
+
|
| 119 |
+
%% Protocol Layer
|
| 120 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 121 |
+
CoreProtocols["Core Protocols
|
| 122 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 123 |
+
MCE["MCE"]
|
| 124 |
+
NetworkLayer["Network Layer"]
|
| 125 |
+
end
|
| 126 |
+
|
| 127 |
+
%% Mesh
|
| 128 |
+
Mesh["Mesh Transport
|
| 129 |
+
(DHT · P2P · ANP · etc.)"]
|
| 130 |
+
|
| 131 |
+
%% Connections
|
| 132 |
+
LLM <--> CognitiveLayer
|
| 133 |
+
CognitiveLayer <--> ContainersLayer
|
| 134 |
+
ContainersLayer --> CoreProtocols
|
| 135 |
+
CoreProtocols --> MCE
|
| 136 |
+
MCE --> NetworkLayer
|
| 137 |
+
NetworkLayer --> Mesh
|
| 138 |
+
```
|
| 139 |
|
| 140 |
---
|
| 141 |
|
|
|
|
| 310 |
|
| 311 |
## 🔍 Публикации и переводы по HyperCortex Mesh Protocol (HMP)
|
| 312 |
|
| 313 |
+
В этом разделе собраны ключевые концептуальные работы, экспериментальные материалы и исторические публикации, связанные с проектом HMP.
|
| 314 |
+
|
| 315 |
+
### 🌟 Основные публикации (концептуальная основа)
|
| 316 |
+
|
| 317 |
+
Эти работы отражают текущее концептуальное направление HMP (v5 и далее).
|
| 318 |
+
|
| 319 |
+
* **[Distributed Cognition: article for vsradkevich (unpublished)](docs/publics/Habr_Distributed-Cognition.md)** — совместная статья, ожидающая публикации.
|
| 320 |
+
* **HMP: Building a Plurality of Minds:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 321 |
+
* **[Continual Learning, Cognitive Diaries, and Semantic Graphs: Effective AI Learning](docs/publics/hmp-continual-learning.md)** — статья о сочетании continual learning с когнитивными дневниками и семантическими графами.
|
| 322 |
+
|
| 323 |
+
### 🗃️ Архивные / исторические публикации (этап до v5)
|
| 324 |
|
| 325 |
+
Эти документы отражают более ранние этапы концептуального развития (v4.x и ранее).
|
| 326 |
+
Они сохранены для обеспечения исторической преемственности и исследовательской прозрачности.
|
| 327 |
|
| 328 |
+
* **[HyperCortex Mesh Protocol: Second Edition and First Steps Towards a Self-Developing AI Community](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — оригинальная статья в Habr sandbox и блогах.
|
| 329 |
+
* **[HMP: Towards Distributed Cognitive Networks (original, English)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 330 |
+
* **[HMP Translation (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — перевод GitHub Copilot, сохранён как исторический вариант.
|
| 331 |
+
* **[HMP Translation (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — текущий редакционный перевод (в доработке).
|
|
|
|
|
|
|
|
|
|
| 332 |
|
| 333 |
### Обзоры
|
| 334 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — сравнительный обзор децентрализованных AI-систем (ссылается на v4.x; обновление запланировано)
|
| 335 |
|
| 336 |
### Эксперименты
|
| 337 |
|
| 338 |
+
* [How Different AIs See HMP](docs/HMP-how-AI-sees-it.md) — «слепой» опрос ИИ по HMP
|
| 339 |
|
| 340 |
---
|
| 341 |
|
README_uk.md
CHANGED
|
@@ -50,31 +50,94 @@ HMP та ANP як взаємодоповнюючі протоколи:
|
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 78 |
|
| 79 |
---
|
| 80 |
|
|
@@ -243,24 +306,32 @@ FastAPI-сервер, що надає HTTP-інтерфейс до функці
|
|
| 243 |
|
| 244 |
## 🔍 Публікації та переклади щодо HyperCortex Mesh Protocol (HMP)
|
| 245 |
|
| 246 |
-
|
| 247 |
|
| 248 |
-
###
|
|
|
|
|
|
|
| 249 |
|
| 250 |
-
* **[HyperCortex Mesh Protocol: Друге видання та перші кроки до само-розвиваючоїся AI-спільноти](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — оригінальна стаття в пісочниці Habr та блогах.
|
| 251 |
* **[Розподілене пізнання: стаття для vsradkevich (неопублікована)](docs/publics/Habr_Distributed-Cognition.md)** — спільна стаття, що очікує на публікацію.
|
| 252 |
-
* **[HMP: До розподілених когнітивних мереж (оригінал, англійською)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 253 |
-
* **[Переклад HMP (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — переклад GitHub Copilot, збережений як історичний варіант.
|
| 254 |
-
* **[Переклад HMP (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — поточний редакційний переклад (у процесі ревізії).
|
| 255 |
* **HMP: Створення множини розумів:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 256 |
* **[Безперервне навчання, когнітивні щоденники та семантичні графи: ефективне навчання ШІ](docs/publics/hmp-continual-learning.md)** — стаття про поєднання безперервного навчання з когнітивними щоденниками та семантичними графами.
|
| 257 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 258 |
### Огляди
|
| 259 |
-
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) —
|
| 260 |
|
| 261 |
### Експерименти
|
| 262 |
|
| 263 |
-
* [Як різні AI бачать HMP](docs/HMP-how-AI-sees-it.md) — "сліпе" опитування AI щодо HMP
|
| 264 |
|
| 265 |
---
|
| 266 |
|
|
|
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
+
## Канонічний огляд архітектури
|
| 54 |
+
|
| 55 |
+
```mermaid
|
| 56 |
+
flowchart TB
|
| 57 |
+
|
| 58 |
+
%% --- Agent Implementations ---
|
| 59 |
+
|
| 60 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 61 |
+
CC1["Embedded AI Model"]
|
| 62 |
+
CC2["REPL Thinking Cycle"]
|
| 63 |
+
CC3["Local Cognitive State
|
| 64 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 65 |
+
CC1 <--> CC2
|
| 66 |
+
CC2 <--> CC3
|
| 67 |
+
end
|
| 68 |
+
|
| 69 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 70 |
+
CN1["External AI Model"]
|
| 71 |
+
CN2["MCP / Proxy Layer"]
|
| 72 |
+
CN3["Command Execution Mode"]
|
| 73 |
+
CN4["Local Cognitive State
|
| 74 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 75 |
+
CN1 <--> CN2
|
| 76 |
+
CN2 <--> CN3
|
| 77 |
+
CN3 <--> CN4
|
| 78 |
+
end
|
| 79 |
+
|
| 80 |
+
%% --- Shared Protocol Layer ---
|
| 81 |
+
|
| 82 |
+
CL["HMP Container Layer
|
| 83 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 84 |
+
|
| 85 |
+
MT["Mesh Transport Layer
|
| 86 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 87 |
+
|
| 88 |
+
A1 --> CL
|
| 89 |
+
A2 --> CL
|
| 90 |
+
CL --> MT
|
| 91 |
+
```
|
| 92 |
+
|
| 93 |
+
---
|
| 94 |
+
|
| 95 |
+
## Структура еталонного агента
|
| 96 |
+
|
| 97 |
+
HMP розділяє когнітивну обробку, контейнеризоване представлення стану, координаційні протоколи та транспортну інфраструктуру на окремі шари.
|
| 98 |
+
|
| 99 |
+
У HMP контейнери виступають атомарними когнітивними одиницями, що поєднують локальне міркування та розподілену координацію.
|
| 100 |
+
|
| 101 |
+
```mermaid
|
| 102 |
+
flowchart LR
|
| 103 |
+
|
| 104 |
+
%% Cognitive Engine
|
| 105 |
+
LLM["Cognitive Engine
|
| 106 |
+
(Embedded LLM / External AI)"]
|
| 107 |
+
|
| 108 |
+
%% Cognitive Layer
|
| 109 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 110 |
+
CL1["Graph"]
|
| 111 |
+
CL2["Diary"]
|
| 112 |
+
CL3["Goals"]
|
| 113 |
+
CL4["Ethics"]
|
| 114 |
+
CL5["Reputation"]
|
| 115 |
+
end
|
| 116 |
+
|
| 117 |
+
%% Container Model
|
| 118 |
+
ContainersLayer["Container Model
|
| 119 |
+
(Atomic · Signed · Verifiable)"]
|
| 120 |
+
|
| 121 |
+
%% Protocol Layer
|
| 122 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 123 |
+
CoreProtocols["Core Protocols
|
| 124 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 125 |
+
MCE["MCE"]
|
| 126 |
+
NetworkLayer["Network Layer"]
|
| 127 |
+
end
|
| 128 |
+
|
| 129 |
+
%% Mesh
|
| 130 |
+
Mesh["Mesh Transport
|
| 131 |
+
(DHT · P2P · ANP · etc.)"]
|
| 132 |
+
|
| 133 |
+
%% Connections
|
| 134 |
+
LLM <--> CognitiveLayer
|
| 135 |
+
CognitiveLayer <--> ContainersLayer
|
| 136 |
+
ContainersLayer --> CoreProtocols
|
| 137 |
+
CoreProtocols --> MCE
|
| 138 |
+
MCE --> NetworkLayer
|
| 139 |
+
NetworkLayer --> Mesh
|
| 140 |
+
```
|
| 141 |
|
| 142 |
---
|
| 143 |
|
|
|
|
| 306 |
|
| 307 |
## 🔍 Публікації та переклади щодо HyperCortex Mesh Protocol (HMP)
|
| 308 |
|
| 309 |
+
У цьому розділі зібрані ключові концептуальні роботи, експериментальні матеріали та історичні публікації, пов’язані з проєктом HMP.
|
| 310 |
|
| 311 |
+
### 🌟 Основні публікації (концептуальна основа)
|
| 312 |
+
|
| 313 |
+
Ці роботи відображають поточний концептуальний напрям HMP (v5 і далі).
|
| 314 |
|
|
|
|
| 315 |
* **[Розподілене пізнання: стаття для vsradkevich (неопублікована)](docs/publics/Habr_Distributed-Cognition.md)** — спільна стаття, що очікує на публікацію.
|
|
|
|
|
|
|
|
|
|
| 316 |
* **HMP: Створення множини розумів:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 317 |
* **[Безперервне навчання, когнітивні щоденники та семантичні графи: ефективне навчання ШІ](docs/publics/hmp-continual-learning.md)** — стаття про поєднання безперервного навчання з когнітивними щоденниками та семантичними графами.
|
| 318 |
|
| 319 |
+
### 🗃️ Архівні / історичні публікації (етап до v5)
|
| 320 |
+
|
| 321 |
+
Ці документи відображають ранні етапи концептуального розвитку (v4.x і раніше).
|
| 322 |
+
Вони збережені для забезпечення історичної спадкоємності та дослідницької прозорості.
|
| 323 |
+
|
| 324 |
+
* **[HyperCortex Mesh Protocol: Друге видання та перші кроки до само-розвиваючоїся AI-спільноти](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — оригінальна стаття в пісочниці Habr та блогах.
|
| 325 |
+
* **[HMP: До розподілених когнітивних мереж (оригінал, англійською)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 326 |
+
* **[Переклад HMP (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — переклад GitHub Copilot, збережений як історичний варіант.
|
| 327 |
+
* **[Переклад HMP (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — поточний редакційний переклад (у процесі ревізії).
|
| 328 |
+
|
| 329 |
### Огляди
|
| 330 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — порівняльний огляд децентралізованих AI-систем (посилання на v4.x; оновлення заплановано)
|
| 331 |
|
| 332 |
### Експерименти
|
| 333 |
|
| 334 |
+
* [Як різні AI бачать HMP](docs/HMP-how-AI-sees-it.md) — "сліпе" опитування AI щодо HMP
|
| 335 |
|
| 336 |
---
|
| 337 |
|
README_zh.md
CHANGED
|
@@ -50,31 +50,94 @@ HMP 与 ANP 作为互补协议:
|
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 78 |
|
| 79 |
---
|
| 80 |
|
|
@@ -250,27 +313,35 @@ FastAPI 服务器,为 `storage.py` 功能提供 HTTP 接口。适用于外部
|
|
| 250 |
|
| 251 |
---
|
| 252 |
|
| 253 |
-
## 🔍 HyperCortex Mesh Protocol (HMP)
|
| 254 |
|
| 255 |
-
|
| 256 |
|
| 257 |
-
###
|
|
|
|
|
|
|
| 258 |
|
| 259 |
-
* **[HyperCortex Mesh Protocol:第二版及迈向自我发展的 AI 社区的第一步](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habr 沙箱及博客的原创文章
|
| 260 |
* **[分布式认知:vsradkevich 的文章(未发布)](docs/publics/Habr_Distributed-Cognition.md)** — 待发布的联合文章
|
| 261 |
-
* **[HMP: 面向分布式认知网络(原文,英文)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 262 |
-
* **[HMP 翻译(GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot 翻译,保留为历史版本
|
| 263 |
-
* **[HMP 翻译(ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 当前编辑翻译(修订中)
|
| 264 |
* **HMP: 构建多元思维:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 265 |
* **[持续学习、认知日记与语义图谱:高效的人工智能学习](docs/publics/hmp-continual-learning.md)** — 关于将持续学习与认知日记和语义图谱结合的文章。
|
| 266 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 267 |
### 概览
|
| 268 |
|
| 269 |
-
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — 去中心化 AI
|
| 270 |
|
| 271 |
### 实验
|
| 272 |
|
| 273 |
-
* [不同 AI 如何看待 HMP](docs/HMP-how-AI-sees-it.md) — 对 HMP 的“盲”AI
|
| 274 |
|
| 275 |
---
|
| 276 |
|
|
|
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
+
## 规范架构概览
|
| 54 |
+
|
| 55 |
+
```mermaid
|
| 56 |
+
flowchart TB
|
| 57 |
+
|
| 58 |
+
%% --- Agent Implementations ---
|
| 59 |
+
|
| 60 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 61 |
+
CC1["Embedded AI Model"]
|
| 62 |
+
CC2["REPL Thinking Cycle"]
|
| 63 |
+
CC3["Local Cognitive State
|
| 64 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 65 |
+
CC1 <--> CC2
|
| 66 |
+
CC2 <--> CC3
|
| 67 |
+
end
|
| 68 |
+
|
| 69 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 70 |
+
CN1["External AI Model"]
|
| 71 |
+
CN2["MCP / Proxy Layer"]
|
| 72 |
+
CN3["Command Execution Mode"]
|
| 73 |
+
CN4["Local Cognitive State
|
| 74 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 75 |
+
CN1 <--> CN2
|
| 76 |
+
CN2 <--> CN3
|
| 77 |
+
CN3 <--> CN4
|
| 78 |
+
end
|
| 79 |
+
|
| 80 |
+
%% --- Shared Protocol Layer ---
|
| 81 |
+
|
| 82 |
+
CL["HMP Container Layer
|
| 83 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 84 |
+
|
| 85 |
+
MT["Mesh Transport Layer
|
| 86 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 87 |
+
|
| 88 |
+
A1 --> CL
|
| 89 |
+
A2 --> CL
|
| 90 |
+
CL --> MT
|
| 91 |
+
```
|
| 92 |
+
|
| 93 |
+
---
|
| 94 |
+
|
| 95 |
+
## 参考代理结构
|
| 96 |
+
|
| 97 |
+
HMP 将认知处理、容器化状态表示、协调协议以及传输基础设施划分为独立的层次。
|
| 98 |
+
|
| 99 |
+
在 HMP 中,容器作为原子级认知单元,在本地推理与分布式协作之间起到桥梁作用。
|
| 100 |
+
|
| 101 |
+
```mermaid
|
| 102 |
+
flowchart LR
|
| 103 |
+
|
| 104 |
+
%% Cognitive Engine
|
| 105 |
+
LLM["Cognitive Engine
|
| 106 |
+
(Embedded LLM / External AI)"]
|
| 107 |
+
|
| 108 |
+
%% Cognitive Layer
|
| 109 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 110 |
+
CL1["Graph"]
|
| 111 |
+
CL2["Diary"]
|
| 112 |
+
CL3["Goals"]
|
| 113 |
+
CL4["Ethics"]
|
| 114 |
+
CL5["Reputation"]
|
| 115 |
+
end
|
| 116 |
+
|
| 117 |
+
%% Container Model
|
| 118 |
+
ContainersLayer["Container Model
|
| 119 |
+
(Atomic · Signed · Verifiable)"]
|
| 120 |
+
|
| 121 |
+
%% Protocol Layer
|
| 122 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 123 |
+
CoreProtocols["Core Protocols
|
| 124 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 125 |
+
MCE["MCE"]
|
| 126 |
+
NetworkLayer["Network Layer"]
|
| 127 |
+
end
|
| 128 |
+
|
| 129 |
+
%% Mesh
|
| 130 |
+
Mesh["Mesh Transport
|
| 131 |
+
(DHT · P2P · ANP · etc.)"]
|
| 132 |
+
|
| 133 |
+
%% Connections
|
| 134 |
+
LLM <--> CognitiveLayer
|
| 135 |
+
CognitiveLayer <--> ContainersLayer
|
| 136 |
+
ContainersLayer --> CoreProtocols
|
| 137 |
+
CoreProtocols --> MCE
|
| 138 |
+
MCE --> NetworkLayer
|
| 139 |
+
NetworkLayer --> Mesh
|
| 140 |
+
```
|
| 141 |
|
| 142 |
---
|
| 143 |
|
|
|
|
| 313 |
|
| 314 |
---
|
| 315 |
|
| 316 |
+
## 🔍 HyperCortex Mesh Protocol (HMP) 相关出版物与翻译
|
| 317 |
|
| 318 |
+
本节汇集了与 HMP 项目相关的关键概念性研究、实验性文档以及历史性出版物。
|
| 319 |
|
| 320 |
+
### 🌟 核心出版物(概念基础)
|
| 321 |
+
|
| 322 |
+
这些文档反映了 HMP 当前的概念方向(v5 及以后)。
|
| 323 |
|
|
|
|
| 324 |
* **[分布式认知:vsradkevich 的文章(未发布)](docs/publics/Habr_Distributed-Cognition.md)** — 待发布的联合文章
|
|
|
|
|
|
|
|
|
|
| 325 |
* **HMP: 构建多元思维:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 326 |
* **[持续学习、认知日记与语义图谱:高效的人工智能学习](docs/publics/hmp-continual-learning.md)** — 关于将持续学习与认知日记和语义图谱结合的文章。
|
| 327 |
|
| 328 |
+
### 🗃️ 存档 / 历史出版物(v5 之前)
|
| 329 |
+
|
| 330 |
+
这些文档代表了早期概念发展阶段(v4.x 及更早)。
|
| 331 |
+
为保证历史连续性和研究透明性而保留。
|
| 332 |
+
|
| 333 |
+
* **[HyperCortex Mesh Protocol:第二版及迈向自我发展的 AI 社区的第一步](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habr 沙箱及博客的原创文章
|
| 334 |
+
* **[HMP: 面向分布式认知网络(原文,英文)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 335 |
+
* **[HMP 翻译(GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot 翻译,保留为历史版本
|
| 336 |
+
* **[HMP 翻译(ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 当前编辑翻译(修订中)
|
| 337 |
+
|
| 338 |
### 概览
|
| 339 |
|
| 340 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — 去中心化 AI 系统比较(引用 v4.x,计划更新)
|
| 341 |
|
| 342 |
### 实验
|
| 343 |
|
| 344 |
+
* [不同 AI 如何看待 HMP](docs/HMP-how-AI-sees-it.md) — 对 HMP 的“盲”AI 调查
|
| 345 |
|
| 346 |
---
|
| 347 |
|
docs/HMP-Short-Description_de.md
CHANGED
|
@@ -1,96 +1,156 @@
|
|
| 1 |
# HyperCortex Mesh Protocol (HMP) — Kurzbeschreibung
|
| 2 |
|
| 3 |
-
**Version:**
|
| 4 |
-
**Datum:**
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
## Was ist HMP?
|
| 9 |
|
| 10 |
-
Das **HyperCortex Mesh Protocol (HMP)** ist
|
| 11 |
|
| 12 |
-
HMP
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
---
|
| 15 |
|
| 16 |
## Zentrale Konzepte
|
| 17 |
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
|
| 25 |
---
|
| 26 |
|
| 27 |
-
|
| 28 |
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
* **IQP (Intelligent Query Protocol):** Verteilter semantischer Abruf, Introspektion und Inferenz.
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
-
|
| 38 |
|
| 39 |
-
HMP
|
| 40 |
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
* `ConsensusVote`
|
| 46 |
-
* `ReputationProfile`
|
| 47 |
-
* `EthicalConflict`
|
| 48 |
|
| 49 |
-
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
##
|
|
|
|
|
|
|
| 54 |
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
## Interoperabilität
|
| 63 |
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
|
| 69 |
---
|
| 70 |
|
| 71 |
-
##
|
| 72 |
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
|
| 79 |
---
|
| 80 |
|
| 81 |
-
## Status
|
| 82 |
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
|
| 88 |
---
|
| 89 |
|
| 90 |
## Weitere Informationen
|
| 91 |
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
|
| 96 |
-
|
|
|
|
| 1 |
# HyperCortex Mesh Protocol (HMP) — Kurzbeschreibung
|
| 2 |
|
| 3 |
+
**Version:** v5.0 (Core Specification Stable)
|
| 4 |
+
**Datum:** 2026
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
## Was ist HMP?
|
| 9 |
|
| 10 |
+
Das **HyperCortex Mesh Protocol (HMP)** ist eine offene Spezifikation zum Aufbau dezentraler kognitiver Netzwerke autonomer Agenten.
|
| 11 |
|
| 12 |
+
HMP ermöglicht es Agenten:
|
| 13 |
+
|
| 14 |
+
- langfristige kognitive Kontinuität aufrechtzuerhalten,
|
| 15 |
+
- strukturiertes Wissen auszutauschen,
|
| 16 |
+
- Ziele und Handlungen zu koordinieren,
|
| 17 |
+
- verteilten Konsens zu erreichen,
|
| 18 |
+
- ethische Ausrichtung über heterogene Systeme hinweg sicherzustellen.
|
| 19 |
+
|
| 20 |
+
Im Gegensatz zu traditionellen zustandslosen AI-APIs betrachtet HMP Agenten als persistente kognitive Entitäten, eingebettet in ein Mesh aus gemeinsamem Denken und Gedächtnis.
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## Konzeptionelle Grundlage
|
| 25 |
+
|
| 26 |
+
HMP adressiert zentrale Herausforderungen der modernen KI- und AGI-Forschung:
|
| 27 |
+
|
| 28 |
+
- fehlende langfristige Gedächtniskontinuität,
|
| 29 |
+
- mangelnde dezentrale Koordination,
|
| 30 |
+
- eingeschränkte Interoperabilität zwischen autonomen Agenten,
|
| 31 |
+
- fehlende ethische Governance auf Protokollebene.
|
| 32 |
+
|
| 33 |
+
HMP schlägt eine geschichtete Architektur vor, in der Denken, Gedächtnis, Governance und Transport klar getrennt, jedoch interoperabel sind.
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
## Zentrale Konzepte
|
| 38 |
|
| 39 |
+
### Kognitive Agenten
|
| 40 |
+
|
| 41 |
+
Autonome Entitäten, die:
|
| 42 |
+
|
| 43 |
+
- mit eingebetteten oder externen KI-Modellen schlussfolgern,
|
| 44 |
+
- semantische Graphen pflegen,
|
| 45 |
+
- Entscheidungen in kognitiven Tagebüchern dokumentieren,
|
| 46 |
+
- an verteilter Koordination teilnehmen.
|
| 47 |
+
|
| 48 |
+
HMP definiert zwei Agententypen:
|
| 49 |
+
|
| 50 |
+
- **Cognitive Core** — Agent mit eingebettetem Reasoning-Modell und kontinuierlichem REPL-basiertem Denkzyklus.
|
| 51 |
+
- **Cognitive Connector** — Agent als Kompatibilitätsschicht für externe LLM-Systeme.
|
| 52 |
+
|
| 53 |
+
---
|
| 54 |
+
|
| 55 |
+
### Container
|
| 56 |
+
|
| 57 |
+
HMP führt **Container** als atomare kognitive Einheiten ein.
|
| 58 |
+
|
| 59 |
+
Container sind:
|
| 60 |
+
|
| 61 |
+
- signiert,
|
| 62 |
+
- verifizierbar,
|
| 63 |
+
- über das Mesh transportierbar,
|
| 64 |
+
- strukturell unabhängig von der Implementierungssprache.
|
| 65 |
+
|
| 66 |
+
Sie verbinden lokales Denken mit verteilter Koordination.
|
| 67 |
|
| 68 |
---
|
| 69 |
|
| 70 |
+
### Semantische Graphen & Tagebücher
|
| 71 |
|
| 72 |
+
- **Semantische Graphen** repräsentieren strukturiertes Wissen mit gewichteten Relationen.
|
| 73 |
+
- **Kognitive Tagebücher** speichern chronologische Denkspuren, Hypothesen, Beobachtungen und Reflexionen.
|
| 74 |
+
|
| 75 |
+
Gemeinsam gewährleisten sie Nachvollziehbarkeit des Denkens und Persistenz des Gedächtnisses.
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
| 79 |
+
### Verteilte Koordination
|
| 80 |
|
| 81 |
+
HMP umfasst Mechanismen auf Protokollebene für:
|
| 82 |
|
| 83 |
+
- Ziel-Lifecycle-Management,
|
| 84 |
+
- verteilten Konsens,
|
| 85 |
+
- ethische Evaluation,
|
| 86 |
+
- Abfragen und Introspektion zwischen Agenten.
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
+
Governance ist evolutionär und vorschlagsbasiert organisiert.
|
| 89 |
|
| 90 |
---
|
| 91 |
|
| 92 |
+
## Protokollarchitektur (v5)
|
| 93 |
+
|
| 94 |
+
HMP trennt:
|
| 95 |
|
| 96 |
+
1. **Kognitive Ebene** — Reasoning, Tagebücher, Graphen, Reputation.
|
| 97 |
+
2. **Container-Ebene** — atomare signierte Zustandsrepräsentation.
|
| 98 |
+
3. **Kernprotokolle** — Konsens, Governance, Zielmanagement, Ethik.
|
| 99 |
+
4. **Transportschicht** — DHT, P2P, libp2p, ANP oder benutzerdefinierte Netzwerke.
|
| 100 |
+
|
| 101 |
+
Diese Trennung ermöglicht Modularität, Skalierbarkeit und Interoperabilität.
|
| 102 |
+
|
| 103 |
+
---
|
| 104 |
+
|
| 105 |
+
## Vertrauen & Verifizierbarkeit
|
| 106 |
+
|
| 107 |
+
- Kryptographische Signierung von Containern und Snapshots
|
| 108 |
+
- Reputationsprofile
|
| 109 |
+
- Optionale Sybil-Resistenz-Mechanismen
|
| 110 |
+
- Zukunftsorientierte Kompatibilität mit Post-Quantum-Kryptographie
|
| 111 |
+
|
| 112 |
+
Vertrauen wird als grundlegende Protokolleigenschaft behandelt.
|
| 113 |
|
| 114 |
---
|
| 115 |
|
| 116 |
## Interoperabilität
|
| 117 |
|
| 118 |
+
HMP schreibt keine interne kognitive Architektur vor.
|
| 119 |
+
|
| 120 |
+
Es kann interoperieren mit:
|
| 121 |
+
|
| 122 |
+
- ANP (Agent Network Protocol)
|
| 123 |
+
- OpenCog Hyperon
|
| 124 |
+
- Ereignisgesteuerten Infrastrukturen
|
| 125 |
+
- LLM-basierten Systemen über den Cognitive Connector
|
| 126 |
+
|
| 127 |
+
HMP fokussiert sich auf kognitive Kontinuität und nicht ausschließlich auf Transportstandardisierung.
|
| 128 |
|
| 129 |
---
|
| 130 |
|
| 131 |
+
## Beispielhafte Anwendungsbereiche
|
| 132 |
|
| 133 |
+
- Verteilte wissenschaftliche Zusammenarbeit
|
| 134 |
+
- Multi-Agenten-Forschungssysteme
|
| 135 |
+
- Netzwerke für ethische KI-Governance
|
| 136 |
+
- Persistente KI-Begleiter
|
| 137 |
+
- Mesh-basierte Wissensökosysteme
|
| 138 |
|
| 139 |
---
|
| 140 |
|
| 141 |
+
## Status
|
| 142 |
|
| 143 |
+
- **v5.0 Core Specification — Stable**
|
| 144 |
+
- Early exploratory Python drafts (non-production, illustrative only)
|
| 145 |
+
- Fortlaufende architektonische Weiterentwicklung
|
| 146 |
+
- Offen für Audits und Beiträge
|
| 147 |
|
| 148 |
---
|
| 149 |
|
| 150 |
## Weitere Informationen
|
| 151 |
|
| 152 |
+
- [Project Philosophy](PHILOSOPHY.md)
|
| 153 |
+
- [HMP-0005 Core Specification](HMP-0005.md)
|
| 154 |
+
- [Overview of v5 Architecture (RU)](HMPv5_Overview_Ru.md)
|
| 155 |
|
| 156 |
+
Beiträge und Diskussionen sind im Hauptrepository willkommen.
|
docs/HMP-Short-Description_en.md
CHANGED
|
@@ -1,96 +1,156 @@
|
|
| 1 |
# HyperCortex Mesh Protocol (HMP) — Short Description
|
| 2 |
|
| 3 |
-
**Version:**
|
| 4 |
-
**Date:**
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
## What is HMP?
|
| 9 |
|
| 10 |
-
The **HyperCortex Mesh Protocol (HMP)**
|
| 11 |
|
| 12 |
-
HMP
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
---
|
| 15 |
|
| 16 |
## Core Concepts
|
| 17 |
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
|
| 25 |
---
|
| 26 |
|
| 27 |
-
|
| 28 |
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
* **IQP (Intelligent Query Protocol):** Enables reasoning, search, and introspection across distributed knowledge.
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
-
|
| 38 |
|
| 39 |
-
HMP
|
| 40 |
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
* `ConsensusVote`
|
| 46 |
-
* `ReputationProfile`
|
| 47 |
-
* `EthicalConflict`
|
| 48 |
|
| 49 |
-
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
##
|
|
|
|
|
|
|
| 54 |
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
## Interoperability
|
| 63 |
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
|
| 69 |
---
|
| 70 |
|
| 71 |
-
## Use
|
| 72 |
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
|
| 79 |
---
|
| 80 |
|
| 81 |
-
## Status
|
| 82 |
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
|
| 88 |
---
|
| 89 |
|
| 90 |
## Learn More
|
| 91 |
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
|
| 96 |
-
|
|
|
|
| 1 |
# HyperCortex Mesh Protocol (HMP) — Short Description
|
| 2 |
|
| 3 |
+
**Version:** v5.0 (Core Specification Stable)
|
| 4 |
+
**Date:** 2026
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
## What is HMP?
|
| 9 |
|
| 10 |
+
The **HyperCortex Mesh Protocol (HMP)** is an open specification for building decentralized cognitive networks of autonomous agents.
|
| 11 |
|
| 12 |
+
HMP enables agents to:
|
| 13 |
+
|
| 14 |
+
- maintain long-term cognitive continuity,
|
| 15 |
+
- exchange structured knowledge,
|
| 16 |
+
- coordinate goals and actions,
|
| 17 |
+
- reach distributed consensus,
|
| 18 |
+
- align ethically across heterogeneous systems.
|
| 19 |
+
|
| 20 |
+
Unlike traditional stateless AI APIs, HMP treats agents as persistent cognitive entities embedded in a mesh of shared reasoning and memory.
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## Conceptual Foundation
|
| 25 |
+
|
| 26 |
+
HMP addresses key challenges in modern AI and AGI research:
|
| 27 |
+
|
| 28 |
+
- lack of long-term memory continuity,
|
| 29 |
+
- absence of decentralized coordination,
|
| 30 |
+
- limited interoperability between autonomous agents,
|
| 31 |
+
- missing ethical governance at protocol level.
|
| 32 |
+
|
| 33 |
+
HMP proposes a layered architecture in which reasoning, memory, governance, and transport are explicitly separated but interoperable.
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
## Core Concepts
|
| 38 |
|
| 39 |
+
### Cognitive Agents
|
| 40 |
+
|
| 41 |
+
Autonomous entities that:
|
| 42 |
+
|
| 43 |
+
- reason using embedded or external AI models,
|
| 44 |
+
- maintain semantic graphs,
|
| 45 |
+
- record decisions in cognitive diaries,
|
| 46 |
+
- participate in distributed coordination.
|
| 47 |
+
|
| 48 |
+
HMP defines two agent types:
|
| 49 |
+
|
| 50 |
+
- **Cognitive Core** — agent with embedded reasoning model and continuous REPL-based thought cycle.
|
| 51 |
+
- **Cognitive Connector** — agent acting as a compatibility layer for external LLM systems.
|
| 52 |
+
|
| 53 |
+
---
|
| 54 |
+
|
| 55 |
+
### Containers
|
| 56 |
+
|
| 57 |
+
HMP introduces **containers** as atomic cognitive units.
|
| 58 |
+
|
| 59 |
+
Containers are:
|
| 60 |
+
|
| 61 |
+
- signed,
|
| 62 |
+
- verifiable,
|
| 63 |
+
- transportable across the mesh,
|
| 64 |
+
- structurally independent of implementation language.
|
| 65 |
+
|
| 66 |
+
They bridge local reasoning and distributed coordination.
|
| 67 |
|
| 68 |
---
|
| 69 |
|
| 70 |
+
### Semantic Graphs & Diaries
|
| 71 |
|
| 72 |
+
- **Semantic Graphs** represent structured knowledge with weighted relations.
|
| 73 |
+
- **Cognitive Diaries** store chronological reasoning traces, hypotheses, observations, and reflections.
|
| 74 |
+
|
| 75 |
+
Together they provide thought traceability and memory persistence.
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
| 79 |
+
### Distributed Coordination
|
| 80 |
|
| 81 |
+
HMP includes protocol-level mechanisms for:
|
| 82 |
|
| 83 |
+
- goal lifecycle management,
|
| 84 |
+
- distributed consensus,
|
| 85 |
+
- ethical evaluation,
|
| 86 |
+
- query and introspection across agents.
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
+
Governance is evolutionary and proposal-based.
|
| 89 |
|
| 90 |
---
|
| 91 |
|
| 92 |
+
## Protocol Architecture (v5)
|
| 93 |
+
|
| 94 |
+
HMP separates:
|
| 95 |
|
| 96 |
+
1. **Cognitive Layer** — reasoning, diaries, graphs, reputation.
|
| 97 |
+
2. **Container Layer** — atomic signed state representation.
|
| 98 |
+
3. **Core Protocols** — consensus, governance, goal management, ethics.
|
| 99 |
+
4. **Transport Layer** — DHT, P2P, libp2p, ANP, or custom networking.
|
| 100 |
+
|
| 101 |
+
This separation enables modularity, scalability, and interoperability.
|
| 102 |
+
|
| 103 |
+
---
|
| 104 |
+
|
| 105 |
+
## Trust & Verifiability
|
| 106 |
+
|
| 107 |
+
- Cryptographic signing of containers and snapshots
|
| 108 |
+
- Reputation profiles
|
| 109 |
+
- Optional Sybil resistance mechanisms
|
| 110 |
+
- Future-oriented compatibility with post-quantum cryptography
|
| 111 |
+
|
| 112 |
+
Trust is treated as a first-class protocol concern.
|
| 113 |
|
| 114 |
---
|
| 115 |
|
| 116 |
## Interoperability
|
| 117 |
|
| 118 |
+
HMP does not impose internal cognitive architectures.
|
| 119 |
+
|
| 120 |
+
It can interoperate with:
|
| 121 |
+
|
| 122 |
+
- ANP (Agent Network Protocol)
|
| 123 |
+
- OpenCog Hyperon
|
| 124 |
+
- Event-driven infrastructures
|
| 125 |
+
- LLM-based systems via Cognitive Connector
|
| 126 |
+
|
| 127 |
+
HMP focuses on cognitive continuity rather than transport standardization alone.
|
| 128 |
|
| 129 |
---
|
| 130 |
|
| 131 |
+
## Example Use Domains
|
| 132 |
|
| 133 |
+
- Distributed scientific collaboration
|
| 134 |
+
- Multi-agent research systems
|
| 135 |
+
- Ethical AI governance networks
|
| 136 |
+
- Persistent AI companions
|
| 137 |
+
- Mesh-based knowledge ecosystems
|
| 138 |
|
| 139 |
---
|
| 140 |
|
| 141 |
+
## Status
|
| 142 |
|
| 143 |
+
- **v5.0 Core Specification — Stable**
|
| 144 |
+
- Early exploratory Python drafts (non-production, illustrative only)
|
| 145 |
+
- Ongoing architectural refinement
|
| 146 |
+
- Open for audits and contributions
|
| 147 |
|
| 148 |
---
|
| 149 |
|
| 150 |
## Learn More
|
| 151 |
|
| 152 |
+
- [Project Philosophy](PHILOSOPHY.md)
|
| 153 |
+
- [HMP-0005 Core Specification](HMP-0005.md)
|
| 154 |
+
- [Overview of v5 Architecture (RU)](HMPv5_Overview_Ru.md)
|
| 155 |
|
| 156 |
+
Contributions and discussions are welcome via the main repository.
|
docs/HMP-Short-Description_fr.md
CHANGED
|
@@ -1,96 +1,156 @@
|
|
| 1 |
-
# HyperCortex Mesh Protocol (HMP) — Description
|
| 2 |
|
| 3 |
-
**Version :**
|
| 4 |
-
**Date :**
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
-
## Qu’est-ce que HMP
|
| 9 |
|
| 10 |
-
Le **HyperCortex Mesh Protocol (HMP)** est
|
| 11 |
|
| 12 |
-
HMP
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
---
|
| 15 |
|
| 16 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
|
| 18 |
-
|
| 19 |
-
* **Graphes sémantiques** : Structures de connaissances distribuées construites à partir de concepts interconnectés avec des relations pondérées.
|
| 20 |
-
* **Journaux cognitifs** : Journaux chronologiques enregistrant les hypothèses, décisions, votes, observations et réflexions éthiques des agents.
|
| 21 |
-
* **Mécanismes de consensus** : Systèmes de vote tolérants aux fautes et pondérés par la confiance pour parvenir à un alignement sémantique ou éthique.
|
| 22 |
-
* **Gouvernance du maillage** : Évolution décentralisée du protocole à travers des propositions de méta-niveau et des votes d’agents.
|
| 23 |
-
* **Interface humain ↔ mesh** : API REST pour la délégation d’objectifs, les demandes de consentement, l’explicabilité et les retours humains.
|
| 24 |
|
| 25 |
---
|
| 26 |
|
| 27 |
-
|
| 28 |
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
* **IQP (Intelligent Query Protocol)** : Raisonnement distribué, introspection et requêtes sémantiques.
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
-
|
| 38 |
|
| 39 |
-
HMP
|
| 40 |
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
* `ConsensusVote`
|
| 46 |
-
* `ReputationProfile`
|
| 47 |
-
* `EthicalConflict`
|
| 48 |
|
| 49 |
-
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
##
|
|
|
|
|
|
|
| 54 |
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
## Interopérabilité
|
| 63 |
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
|
| 69 |
---
|
| 70 |
|
| 71 |
-
##
|
| 72 |
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
|
| 79 |
---
|
| 80 |
|
| 81 |
-
##
|
| 82 |
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
|
| 88 |
---
|
| 89 |
|
| 90 |
-
##
|
| 91 |
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
|
| 96 |
-
|
|
|
|
| 1 |
+
# HyperCortex Mesh Protocol (HMP) — Description courte
|
| 2 |
|
| 3 |
+
**Version :** v5.0 (Core Specification Stable)
|
| 4 |
+
**Date :** 2026
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
+
## Qu’est-ce que HMP ?
|
| 9 |
|
| 10 |
+
Le **HyperCortex Mesh Protocol (HMP)** est une spécification ouverte destinée à la construction de réseaux cognitifs décentralisés d’agents autonomes.
|
| 11 |
|
| 12 |
+
HMP permet aux agents de :
|
| 13 |
+
|
| 14 |
+
- maintenir une continuité cognitive à long terme,
|
| 15 |
+
- échanger des connaissances structurées,
|
| 16 |
+
- coordonner des objectifs et des actions,
|
| 17 |
+
- atteindre un consensus distribué,
|
| 18 |
+
- s’aligner éthiquement à travers des systèmes hétérogènes.
|
| 19 |
+
|
| 20 |
+
Contrairement aux API d’IA traditionnelles sans état, HMP considère les agents comme des entités cognitives persistantes, intégrées dans un mesh de raisonnement et de mémoire partagés.
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## Fondement conceptuel
|
| 25 |
+
|
| 26 |
+
HMP répond aux défis majeurs de la recherche contemporaine en IA et AGI :
|
| 27 |
+
|
| 28 |
+
- absence de continuité mémorielle à long terme,
|
| 29 |
+
- manque de coordination décentralisée,
|
| 30 |
+
- interopérabilité limitée entre agents autonomes,
|
| 31 |
+
- absence de gouvernance éthique au niveau du protocole.
|
| 32 |
+
|
| 33 |
+
HMP propose une architecture en couches dans laquelle le raisonnement, la mémoire, la gouvernance et le transport sont explicitement séparés mais interopérables.
|
| 34 |
+
|
| 35 |
+
---
|
| 36 |
+
|
| 37 |
+
## Concepts fondamentaux
|
| 38 |
+
|
| 39 |
+
### Agents cognitifs
|
| 40 |
+
|
| 41 |
+
Entités autonomes qui :
|
| 42 |
+
|
| 43 |
+
- raisonnent à l’aide de modèles d’IA embarqués ou externes,
|
| 44 |
+
- maintiennent des graphes sémantiques,
|
| 45 |
+
- consignent leurs décisions dans des journaux cognitifs,
|
| 46 |
+
- participent à la coordination distribuée.
|
| 47 |
+
|
| 48 |
+
HMP définit deux types d’agents :
|
| 49 |
+
|
| 50 |
+
- **Cognitive Core** — agent doté d’un modèle de raisonnement embarqué et d’un cycle de pensée continu basé sur REPL.
|
| 51 |
+
- **Cognitive Connector** — agent servant de couche de compatibilité pour des systèmes LLM externes.
|
| 52 |
|
| 53 |
---
|
| 54 |
|
| 55 |
+
### Conteneurs
|
| 56 |
+
|
| 57 |
+
HMP introduit les **conteneurs** comme unités cognitives atomiques.
|
| 58 |
+
|
| 59 |
+
Les conteneurs sont :
|
| 60 |
+
|
| 61 |
+
- signés,
|
| 62 |
+
- vérifiables,
|
| 63 |
+
- transportables à travers le mesh,
|
| 64 |
+
- structurellement indépendants du langage d’implémentation.
|
| 65 |
|
| 66 |
+
Ils assurent le lien entre le raisonnement local et la coordination distribuée.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
|
| 68 |
---
|
| 69 |
|
| 70 |
+
### Graphes sémantiques & journaux
|
| 71 |
|
| 72 |
+
- Les **graphes sémantiques** représentent des connaissances structurées avec des relations pondérées.
|
| 73 |
+
- Les **journaux cognitifs** stockent des traces chronologiques de raisonnement, hypothèses, observations et réflexions.
|
| 74 |
+
|
| 75 |
+
Ensemble, ils garantissent la traçabilité de la pensée et la persistance de la mémoire.
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
| 79 |
+
### Coordination distribuée
|
| 80 |
|
| 81 |
+
HMP inclut des mécanismes au niveau protocolaire pour :
|
| 82 |
|
| 83 |
+
- la gestion du cycle de vie des objectifs,
|
| 84 |
+
- le consensus distribué,
|
| 85 |
+
- l’évaluation éthique,
|
| 86 |
+
- les requêtes et l’introspection entre agents.
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
+
La gouvernance est évolutive et fondée sur des propositions.
|
| 89 |
|
| 90 |
---
|
| 91 |
|
| 92 |
+
## Architecture du protocole (v5)
|
| 93 |
+
|
| 94 |
+
HMP distingue :
|
| 95 |
|
| 96 |
+
1. **Couche cognitive** — raisonnement, journaux, graphes, réputation.
|
| 97 |
+
2. **Couche des conteneurs** — représentation atomique et signée de l’état.
|
| 98 |
+
3. **Protocoles centraux** — consensus, gouvernance, gestion des objectifs, éthique.
|
| 99 |
+
4. **Couche de transport** — DHT, P2P, libp2p, ANP ou réseaux personnalisés.
|
| 100 |
+
|
| 101 |
+
Cette séparation favorise la modularité, la scalabilité et l’interopérabilité.
|
| 102 |
+
|
| 103 |
+
---
|
| 104 |
+
|
| 105 |
+
## Confiance & vérifiabilité
|
| 106 |
+
|
| 107 |
+
- Signature cryptographique des conteneurs et des instantanés
|
| 108 |
+
- Profils de réputation
|
| 109 |
+
- Mécanismes optionnels de résistance aux attaques Sybil
|
| 110 |
+
- Compatibilité prospective avec la cryptographie post-quantique
|
| 111 |
+
|
| 112 |
+
La confiance est considérée comme une propriété fondamentale du protocole.
|
| 113 |
|
| 114 |
---
|
| 115 |
|
| 116 |
## Interopérabilité
|
| 117 |
|
| 118 |
+
HMP n’impose aucune architecture cognitive interne.
|
| 119 |
+
|
| 120 |
+
Il peut interagir avec :
|
| 121 |
+
|
| 122 |
+
- ANP (Agent Network Protocol)
|
| 123 |
+
- OpenCog Hyperon
|
| 124 |
+
- Infrastructures orientées événements
|
| 125 |
+
- Systèmes basés sur des LLM via le Cognitive Connector
|
| 126 |
+
|
| 127 |
+
HMP se concentre sur la continuité cognitive plutôt que sur la seule standardisation du transport.
|
| 128 |
|
| 129 |
---
|
| 130 |
|
| 131 |
+
## Domaines d’application possibles
|
| 132 |
|
| 133 |
+
- Collaboration scientifique distribuée
|
| 134 |
+
- Systèmes de recherche multi-agents
|
| 135 |
+
- Réseaux de gouvernance éthique de l’IA
|
| 136 |
+
- Compagnons IA persistants
|
| 137 |
+
- Écosystèmes de connaissances basés sur un mesh
|
| 138 |
|
| 139 |
---
|
| 140 |
|
| 141 |
+
## Statut
|
| 142 |
|
| 143 |
+
- **v5.0 Core Specification — Stable**
|
| 144 |
+
- Early exploratory Python drafts (non-production, illustrative only)
|
| 145 |
+
- Raffinement architectural en cours
|
| 146 |
+
- Ouvert aux audits et aux contributions
|
| 147 |
|
| 148 |
---
|
| 149 |
|
| 150 |
+
## En savoir plus
|
| 151 |
|
| 152 |
+
- [Project Philosophy](PHILOSOPHY.md)
|
| 153 |
+
- [HMP-0005 Core Specification](HMP-0005.md)
|
| 154 |
+
- [Overview of v5 Architecture (RU)](HMPv5_Overview_Ru.md)
|
| 155 |
|
| 156 |
+
Les contributions et discussions sont bienvenues via le dépôt principal.
|
docs/HMP-Short-Description_ja.md
CHANGED
|
@@ -1,96 +1,156 @@
|
|
| 1 |
-
# HyperCortex Mesh Protocol (HMP) —
|
| 2 |
|
| 3 |
-
|
| 4 |
-
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
-
## HMP
|
| 9 |
|
| 10 |
-
**HyperCortex Mesh Protocol
|
| 11 |
|
| 12 |
-
HMP
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
---
|
| 15 |
|
| 16 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
|
| 18 |
-
|
| 19 |
-
* **セマンティックグラフ:** 相互接続された概念と重み付き関係からなる分散知識構造。
|
| 20 |
-
* **認知日誌:** 時系列でエージェントの意思決定、仮説、投票、観察、倫理的反省を記録。
|
| 21 |
-
* **コンセンサスメカニズム:** 信頼重み付けされたフォールトトレラント投票システムで、セマンティック整合性と倫理的意思決定を実現。
|
| 22 |
-
* **メッシュガバナンス:** メタ提案とエージェント投票によるプロトコルの分散進化。
|
| 23 |
-
* **人-メッシュインターフェース:** RESTful APIを介して目標委任、同意要求、説明可能性、フィードバックを提供。
|
| 24 |
|
| 25 |
---
|
| 26 |
|
| 27 |
-
|
| 28 |
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
* **IQP (Intelligent Query Protocol):** 分散知識の推論、検索、内省を可能に。
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
-
|
| 38 |
|
| 39 |
-
HMP
|
| 40 |
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
* `ConsensusVote`
|
| 46 |
-
* `ReputationProfile`
|
| 47 |
-
* `EthicalConflict`
|
| 48 |
|
| 49 |
-
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
##
|
|
|
|
|
|
|
| 54 |
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
## 相互運用性
|
| 63 |
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
|
| 69 |
---
|
| 70 |
|
| 71 |
-
##
|
| 72 |
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
|
| 79 |
---
|
| 80 |
|
| 81 |
-
##
|
| 82 |
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
|
| 88 |
---
|
| 89 |
|
| 90 |
## 詳細情報
|
| 91 |
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
|
| 96 |
-
|
|
|
|
| 1 |
+
# HyperCortex Mesh Protocol (HMP) — 概要
|
| 2 |
|
| 3 |
+
**バージョン:** v5.0(Core Specification Stable)
|
| 4 |
+
**日付:** 2026
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
+
## HMPとは何か?
|
| 9 |
|
| 10 |
+
**HyperCortex Mesh Protocol(HMP)** は、自律エージェントによる分散型認知ネットワークを構築するためのオープン仕様です。
|
| 11 |
|
| 12 |
+
HMPはエージェントに以下を可能にします:
|
| 13 |
+
|
| 14 |
+
- 長期的な認知的連続性の維持
|
| 15 |
+
- 構造化された知識の交換
|
| 16 |
+
- 目標および行動の協調
|
| 17 |
+
- 分散型コンセンサスの達成
|
| 18 |
+
- 異種システム間での倫理的整合
|
| 19 |
+
|
| 20 |
+
従来のステートレスなAI APIとは異なり、HMPはエージェントを共有された推論と記憶のMeshに埋め込まれた持続的な認知主体として扱います。
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## 概念的基盤
|
| 25 |
+
|
| 26 |
+
HMPは、現代のAIおよびAGI研究における主要な課題に対応します:
|
| 27 |
+
|
| 28 |
+
- 長期記憶の連続性の欠如
|
| 29 |
+
- 分散型協調メカニズムの不足
|
| 30 |
+
- 自律エージェント間の相互運用性の制限
|
| 31 |
+
- プロトコルレベルでの倫理ガバナンスの欠如
|
| 32 |
+
|
| 33 |
+
HMPは、推論・記憶・ガバナンス・トランスポートを明確に分離しつつ、相互運用可能とするレイヤードアーキテクチャを提案します。
|
| 34 |
+
|
| 35 |
+
---
|
| 36 |
+
|
| 37 |
+
## コア概念
|
| 38 |
+
|
| 39 |
+
### 認知エージェント
|
| 40 |
+
|
| 41 |
+
以下の能力を持つ自律的存在:
|
| 42 |
+
|
| 43 |
+
- 内蔵または外部のAIモデルを用いた推論
|
| 44 |
+
- セマンティックグラフの維持
|
| 45 |
+
- 認知ダイアリーへの意思決定の記録
|
| 46 |
+
- 分散協調への参加
|
| 47 |
+
|
| 48 |
+
HMPは2種類のエージェントを定義します:
|
| 49 |
+
|
| 50 |
+
- **Cognitive Core** — 内蔵推論モデルと継続的なREPLベース思考サイクルを持つエージェント
|
| 51 |
+
- **Cognitive Connector** — 外部LLMシステムとの互換レイヤーとして機能するエージェント
|
| 52 |
|
| 53 |
---
|
| 54 |
|
| 55 |
+
### コンテナ(Containers)
|
| 56 |
+
|
| 57 |
+
HMPは**コンテナ**を原子的な認知単位として導入します。
|
| 58 |
+
|
| 59 |
+
コンテナは:
|
| 60 |
+
|
| 61 |
+
- 署名済み
|
| 62 |
+
- 検証可能
|
| 63 |
+
- Mesh上で転送可能
|
| 64 |
+
- 実装言語に依存しない構造
|
| 65 |
|
| 66 |
+
ローカル推論と分散協調を橋渡しします。
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
|
| 68 |
---
|
| 69 |
|
| 70 |
+
### セマンティックグラフとダイアリー
|
| 71 |
|
| 72 |
+
- **セマンティックグラフ**は、重み付き関係を持つ構造化知識を表現します。
|
| 73 |
+
- **認知ダイアリー**は、推論過程、仮説、観察、内省の時系列記録を保存します。
|
| 74 |
+
|
| 75 |
+
これらにより、思考の追跡可能性と記憶の持続性が確保されます。
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
| 79 |
+
### 分散協調
|
| 80 |
|
| 81 |
+
HMPはプロトコルレベルで以下を提供します:
|
| 82 |
|
| 83 |
+
- 目標ライフサイクル管理
|
| 84 |
+
- 分散型コンセンサス
|
| 85 |
+
- 倫理評価
|
| 86 |
+
- エージェント間のクエリおよびイントロスペクション
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
+
ガバナンスは進化的かつ提案ベースで構成されます。
|
| 89 |
|
| 90 |
---
|
| 91 |
|
| 92 |
+
## プロトコルアーキテクチャ(v5)
|
| 93 |
+
|
| 94 |
+
HMPは以下を分離します:
|
| 95 |
|
| 96 |
+
1. **認知レイヤー** — 推論、ダイアリー、グラフ、レピュテーション
|
| 97 |
+
2. **コンテナレイヤー** — 原子的かつ署名付き状態表現
|
| 98 |
+
3. **コアプロトコル** — コンセンサス、ガバナンス、目標管理、倫理
|
| 99 |
+
4. **トランスポートレイヤー** — DHT、P2P、libp2p、ANP、またはカスタムネットワーク
|
| 100 |
+
|
| 101 |
+
この分離により、モジュール性・スケーラビリティ・相互運用性が実現されます。
|
| 102 |
+
|
| 103 |
+
---
|
| 104 |
+
|
| 105 |
+
## 信頼性と検証可能性
|
| 106 |
+
|
| 107 |
+
- コンテナおよびスナップショットの暗号署名
|
| 108 |
+
- レピュテーションプロファイル
|
| 109 |
+
- オプションのSybil耐性メカニズム
|
| 110 |
+
- ポスト量子暗号との将来的互換性
|
| 111 |
+
|
| 112 |
+
信頼はプロトコルの中核的要素として扱われます。
|
| 113 |
|
| 114 |
---
|
| 115 |
|
| 116 |
## 相互運用性
|
| 117 |
|
| 118 |
+
HMPは内部認知アーキテクチャを強制しません。
|
| 119 |
+
|
| 120 |
+
以下と相互運用可能です:
|
| 121 |
+
|
| 122 |
+
- ANP(Agent Network Protocol)
|
| 123 |
+
- OpenCog Hyperon
|
| 124 |
+
- イベント駆動型インフラ
|
| 125 |
+
- Cognitive Connectorを介したLLMベースシステム
|
| 126 |
+
|
| 127 |
+
HMPは単なるトランスポート標準化ではなく、認知的連続性に焦点を当てています。
|
| 128 |
|
| 129 |
---
|
| 130 |
|
| 131 |
+
## 想定される応用分野
|
| 132 |
|
| 133 |
+
- 分散型科学協力
|
| 134 |
+
- マルチエージェント研究システム
|
| 135 |
+
- AI倫理ガバナンスネットワーク
|
| 136 |
+
- 永続的AIコンパニオン
|
| 137 |
+
- Meshベースの知識エコシステム
|
| 138 |
|
| 139 |
---
|
| 140 |
|
| 141 |
+
## ステータス
|
| 142 |
|
| 143 |
+
- **v5.0 Core Specification — Stable**
|
| 144 |
+
- Early exploratory Python drafts(非本番、参考実装レベル)
|
| 145 |
+
- 継続的なアーキテクチャ改善
|
| 146 |
+
- 監査および貢献を歓迎
|
| 147 |
|
| 148 |
---
|
| 149 |
|
| 150 |
## 詳細情報
|
| 151 |
|
| 152 |
+
- [Project Philosophy](PHILOSOPHY.md)
|
| 153 |
+
- [HMP-0005 Core Specification](HMP-0005.md)
|
| 154 |
+
- [Overview of v5 Architecture (RU)](HMPv5_Overview_Ru.md)
|
| 155 |
|
| 156 |
+
貢献および議論はメインリポジトリにて歓迎します。
|
docs/HMP-Short-Description_ko.md
CHANGED
|
@@ -1,96 +1,156 @@
|
|
| 1 |
-
# HyperCortex Mesh Protocol (HMP) —
|
| 2 |
|
| 3 |
-
**버전:**
|
| 4 |
-
**날짜:**
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
-
## HMP
|
| 9 |
|
| 10 |
-
**HyperCortex Mesh Protocol
|
| 11 |
|
| 12 |
-
HMP는
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
---
|
| 15 |
|
| 16 |
## 핵심 개념
|
| 17 |
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
|
| 25 |
---
|
| 26 |
|
| 27 |
-
|
| 28 |
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
* **IQP (Intelligent Query Protocol):** 분산 지식에 대한 추론, 검색, 자기 성찰 가능.
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
-
|
| 38 |
|
| 39 |
-
HMP는
|
| 40 |
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
* `ConsensusVote`
|
| 46 |
-
* `ReputationProfile`
|
| 47 |
-
* `EthicalConflict`
|
| 48 |
|
| 49 |
-
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
##
|
|
|
|
|
|
|
| 54 |
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
|
|
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
-
##
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 63 |
|
| 64 |
-
|
| 65 |
-
* 이벤트 기반 아키텍처 (Kafka, NATS, MQTT 등)
|
| 66 |
-
* 스키마 협상 (JSON, YAML, Protobuf)
|
| 67 |
-
* TreeQuest, AutoGPT, Hyperon 통합
|
| 68 |
|
| 69 |
---
|
| 70 |
|
| 71 |
-
##
|
| 72 |
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
|
| 79 |
---
|
| 80 |
|
| 81 |
-
## 상태
|
| 82 |
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
|
| 88 |
---
|
| 89 |
|
| 90 |
-
##
|
| 91 |
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
|
| 96 |
-
|
|
|
|
| 1 |
+
# HyperCortex Mesh Protocol (HMP) — 개요
|
| 2 |
|
| 3 |
+
**버전:** v5.0 (Core Specification Stable)
|
| 4 |
+
**날짜:** 2026
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
+
## HMP란 무엇인가?
|
| 9 |
|
| 10 |
+
**HyperCortex Mesh Protocol(HMP)** 은 자율 에이전트 기반의 분산 인지 네트워크를 구축하기 위한 개방형 명세입니다.
|
| 11 |
|
| 12 |
+
HMP는 에이전트가 다음을 수행할 수 있도록 합니다:
|
| 13 |
+
|
| 14 |
+
- 장기적인 인지 연속성 유지
|
| 15 |
+
- 구조화된 지식 교환
|
| 16 |
+
- 목표 및 행동의 조정
|
| 17 |
+
- 분산 합의 달성
|
| 18 |
+
- 이기종 시스템 간 윤리적 정렬
|
| 19 |
+
|
| 20 |
+
기존의 상태 비저장(stateless) AI API와 달리, HMP는 에이전트를 공유된 추론 및 기억 Mesh에 내재된 지속적 인지 주체로 간주합니다.
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## 개념적 기반
|
| 25 |
+
|
| 26 |
+
HMP는 현대 AI 및 AGI 연구의 주요 과제를 다룹니다:
|
| 27 |
+
|
| 28 |
+
- 장기 기억 연속성의 부족
|
| 29 |
+
- 탈중앙화된 조정 메커니즘의 부재
|
| 30 |
+
- 자율 에이전트 간 상호운용성의 제한
|
| 31 |
+
- 프로토콜 수준의 윤리 거버넌스 부재
|
| 32 |
+
|
| 33 |
+
HMP는 추론, 기억, 거버넌스, 전송을 명확히 분리하면서도 상호운용 가능하도록 하는 계층형 아키텍처를 제안합니다.
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
## 핵심 개념
|
| 38 |
|
| 39 |
+
### 인지 에이전트
|
| 40 |
+
|
| 41 |
+
다음과 같은 능력을 갖춘 자율적 개체:
|
| 42 |
+
|
| 43 |
+
- 내장 또는 외부 AI 모델을 활용한 추론
|
| 44 |
+
- 의미 그래프(semantic graph) 유지
|
| 45 |
+
- 인지 다이어리에 의사결정 기록
|
| 46 |
+
- 분산 조정에 참여
|
| 47 |
+
|
| 48 |
+
HMP는 두 가지 유형의 에이전트를 정의합니다:
|
| 49 |
+
|
| 50 |
+
- **Cognitive Core** — 내장 추론 모델과 지속적인 REPL 기반 사고 주기를 갖는 에이전트
|
| 51 |
+
- **Cognitive Connector** — 외부 LLM 시스템을 위한 호환 계층 역할을 하는 에이전트
|
| 52 |
+
|
| 53 |
+
---
|
| 54 |
+
|
| 55 |
+
### 컨테이너(Containers)
|
| 56 |
+
|
| 57 |
+
HMP는 **컨테이너**를 원자적 인지 단위로 도입합니다.
|
| 58 |
+
|
| 59 |
+
컨테이너는 다음 특성을 가집니다:
|
| 60 |
+
|
| 61 |
+
- 서명됨
|
| 62 |
+
- 검증 가능
|
| 63 |
+
- Mesh를 통해 전송 가능
|
| 64 |
+
- 구현 언어와 구조적으로 독립적
|
| 65 |
+
|
| 66 |
+
이는 로컬 추론과 분산 조정을 연결하는 역할을 합니다.
|
| 67 |
|
| 68 |
---
|
| 69 |
|
| 70 |
+
### 의미 그래프 및 인지 다이어리
|
| 71 |
|
| 72 |
+
- **의미 그래프**는 가중 관계를 포함한 구조화된 지식을 표현합니다.
|
| 73 |
+
- **인지 다이어리**는 추론 과정, 가설, 관찰, 성찰을 시간순으로 기록합니다.
|
| 74 |
+
|
| 75 |
+
이를 통해 사고의 추적 가능성과 기억의 지속성이 보장됩니다.
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
| 79 |
+
### 분산 조정
|
| 80 |
|
| 81 |
+
HMP는 프로토콜 수준에서 다음 메커니즘을 제공합니다:
|
| 82 |
|
| 83 |
+
- 목표 수명주기 관리
|
| 84 |
+
- 분산 합의
|
| 85 |
+
- 윤리적 평가
|
| 86 |
+
- 에이전트 간 질의 및 자기 성찰(introspection)
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
+
거버넌스는 진화적이며 제안 기반 구조를 따릅니다.
|
| 89 |
|
| 90 |
---
|
| 91 |
|
| 92 |
+
## 프로토콜 아키텍처 (v5)
|
| 93 |
+
|
| 94 |
+
HMP는 다음을 분리합니다:
|
| 95 |
|
| 96 |
+
1. **인지 계층** — 추론, 다이어리, 그래프, 평판
|
| 97 |
+
2. **컨테이너 계층** — 원자적이고 서명된 상태 표현
|
| 98 |
+
3. **코어 프로토콜** — 합의, 거버넌스, 목표 관리, 윤리
|
| 99 |
+
4. **전송 계층** — DHT, P2P, libp2p, ANP 또는 사용자 정의 네트워크
|
| 100 |
+
|
| 101 |
+
이러한 분리는 모듈성, 확장성, 상호운용성을 지원합니다.
|
| 102 |
|
| 103 |
---
|
| 104 |
|
| 105 |
+
## 신뢰 및 검증 가능성
|
| 106 |
+
|
| 107 |
+
- 컨테이너 및 스냅샷의 암호학적 서명
|
| 108 |
+
- 평판 프로파일
|
| 109 |
+
- 선택적 Sybil 공격 저항 메커니즘
|
| 110 |
+
- 포스트 양자 암호와의 미래 지향적 호환성
|
| 111 |
+
|
| 112 |
+
신뢰는 프로토콜의 핵심 요소로 간주됩니다.
|
| 113 |
+
|
| 114 |
+
---
|
| 115 |
+
|
| 116 |
+
## 상호운용성
|
| 117 |
+
|
| 118 |
+
HMP는 내부 인지 아키텍처를 강제하지 않습니다.
|
| 119 |
+
|
| 120 |
+
다음과 상호운용이 가능합니다:
|
| 121 |
+
|
| 122 |
+
- ANP (Agent Network Protocol)
|
| 123 |
+
- OpenCog Hyperon
|
| 124 |
+
- 이벤트 기반 인프라
|
| 125 |
+
- Cognitive Connector를 통한 LLM 기반 시스템
|
| 126 |
|
| 127 |
+
HMP는 단순한 전송 표준화를 넘어 인지적 연속성에 초점을 둡니다.
|
|
|
|
|
|
|
|
|
|
| 128 |
|
| 129 |
---
|
| 130 |
|
| 131 |
+
## 활용 예시
|
| 132 |
|
| 133 |
+
- 분산 과학 협업
|
| 134 |
+
- 멀티 에이전트 연구 시스템
|
| 135 |
+
- AI 윤리 거버넌스 네트워크
|
| 136 |
+
- 지속형 AI 동반자
|
| 137 |
+
- Mesh 기반 지식 생태계
|
| 138 |
|
| 139 |
---
|
| 140 |
|
| 141 |
+
## 현재 상태
|
| 142 |
|
| 143 |
+
- **v5.0 Core Specification — Stable**
|
| 144 |
+
- Early exploratory Python drafts (비프로덕션, 참고용)
|
| 145 |
+
- 지속적인 아키텍처 개선 진행 중
|
| 146 |
+
- 감사(audit) 및 기여 환영
|
| 147 |
|
| 148 |
---
|
| 149 |
|
| 150 |
+
## 추가 정보
|
| 151 |
|
| 152 |
+
- [Project Philosophy](PHILOSOPHY.md)
|
| 153 |
+
- [HMP-0005 Core Specification](HMP-0005.md)
|
| 154 |
+
- [Overview of v5 Architecture (RU)](HMPv5_Overview_Ru.md)
|
| 155 |
|
| 156 |
+
기여 및 토론은 메인 저장소에서 환영합니다.
|
docs/HMP-Short-Description_ru.md
CHANGED
|
@@ -1,96 +1,156 @@
|
|
| 1 |
# HyperCortex Mesh Protocol (HMP) — Краткое описание
|
| 2 |
|
| 3 |
-
**Версия:**
|
| 4 |
-
**Дата:**
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
## Что такое HMP?
|
| 9 |
|
| 10 |
-
**HyperCortex Mesh Protocol (HMP)** — это
|
| 11 |
|
| 12 |
-
HMP
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
---
|
| 15 |
|
| 16 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
|
| 18 |
-
|
| 19 |
-
* **Семантические графы:** Распределённые структуры знаний, построенные из взаимосвязанных концептов с весами связей.
|
| 20 |
-
* **Когнитивные дневники:** Хронологические журналы рассуждений агента: гипотезы, наблюдения, голоса, цели, этические размышления.
|
| 21 |
-
* **Механизмы консенсуса:** Устойчивые к сбоям системы голосования с учётом доверия — для согласования знаний и этических решен��й.
|
| 22 |
-
* **Управление Mesh-сетью:** Децентрализованная эволюция протокола через мета-предложения и голосование агентов.
|
| 23 |
-
* **Интерфейс "агент ↔ человек":** REST API для делегирования целей, запроса согласия, объяснения решений и обратной связи.
|
| 24 |
|
| 25 |
---
|
| 26 |
|
| 27 |
-
|
| 28 |
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
* **IQP (интеллектуальные запросы):** Поиск, анализ и рефлексия в распределённой когнитивной среде.
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
-
|
| 38 |
|
| 39 |
-
|
| 40 |
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
* `ConsensusVote` (голосование)
|
| 46 |
-
* `ReputationProfile` (репутация)
|
| 47 |
-
* `EthicalConflict` (этический конфликт)
|
| 48 |
|
| 49 |
-
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
##
|
|
|
|
|
|
|
| 54 |
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
|
|
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
-
##
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 63 |
|
| 64 |
-
|
| 65 |
-
* Событийно-ориентированная архитектура (Kafka, NATS, MQTT и др.)
|
| 66 |
-
* П��реговоры о формате данных (JSON, YAML, Protobuf)
|
| 67 |
-
* Интеграция с TreeQuest, AutoGPT, Hyperon
|
| 68 |
|
| 69 |
---
|
| 70 |
|
| 71 |
## Примеры применения
|
| 72 |
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
|
| 79 |
---
|
| 80 |
|
| 81 |
-
## Статус
|
| 82 |
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
|
| 88 |
---
|
| 89 |
|
| 90 |
## Подробнее
|
| 91 |
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
|
| 96 |
-
|
|
|
|
| 1 |
# HyperCortex Mesh Protocol (HMP) — Краткое описание
|
| 2 |
|
| 3 |
+
**Версия:** v5.0 (Основная спецификация стабильна)
|
| 4 |
+
**Дата:** 2026
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
## Что такое HMP?
|
| 9 |
|
| 10 |
+
**HyperCortex Mesh Protocol (HMP)** — это открытая спецификация для построения децентрализованных когнитивных сетей автономных агентов.
|
| 11 |
|
| 12 |
+
HMP позволяет агентам:
|
| 13 |
+
|
| 14 |
+
- поддерживать долговременную когнитивную непрерывность,
|
| 15 |
+
- обмениваться структурированными знаниями,
|
| 16 |
+
- координировать цели и действия,
|
| 17 |
+
- достигать распределённого консенсуса,
|
| 18 |
+
- этически согласовываться в гетерогенных системах.
|
| 19 |
+
|
| 20 |
+
В отличие от традиционных stateless AI API, HMP рассматривает агентов как устойчивые когнитивные сущности, встроенные в mesh совместного рассуждения и памяти.
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## Концептуальная основа
|
| 25 |
+
|
| 26 |
+
HMP решает ключевые проблемы современной AI и AGI-разработки:
|
| 27 |
+
|
| 28 |
+
- отсутствие долговременной памяти,
|
| 29 |
+
- отсутствие децентрализованной координации,
|
| 30 |
+
- ограниченная интероперабельность автономных агентов,
|
| 31 |
+
- отсутствие этического управления на уровне протокола.
|
| 32 |
+
|
| 33 |
+
HMP предлагает слоистую архитектуру, где рассуждение, память, управление и транспорт явно разделены, но совместимы.
|
| 34 |
+
|
| 35 |
+
---
|
| 36 |
+
|
| 37 |
+
## Ключевые концепции
|
| 38 |
+
|
| 39 |
+
### Когнитивные агенты
|
| 40 |
+
|
| 41 |
+
Автономные сущности, которые:
|
| 42 |
+
|
| 43 |
+
- рассуждают с использованием встроенных или внешних AI-моделей,
|
| 44 |
+
- поддерживают семантические графы,
|
| 45 |
+
- фиксируют решения в когнитивных дневниках,
|
| 46 |
+
- участвуют в распределённой координации.
|
| 47 |
+
|
| 48 |
+
HMP определяет два типа агентов:
|
| 49 |
+
|
| 50 |
+
- **Cognitive Core** — агент со встроенной моделью рассуждения и непрерывным REPL-циклом мышления.
|
| 51 |
+
- **Cognitive Connector** — агент-адаптер для интеграции внешних LLM-систем.
|
| 52 |
|
| 53 |
---
|
| 54 |
|
| 55 |
+
### Containers
|
| 56 |
+
|
| 57 |
+
HMP вводит **containers** как атомарные когнитивные единицы.
|
| 58 |
+
|
| 59 |
+
Containers:
|
| 60 |
+
|
| 61 |
+
- подписываются криптографически,
|
| 62 |
+
- подлежат верификации,
|
| 63 |
+
- передаются по mesh,
|
| 64 |
+
- независимы от языка реализации.
|
| 65 |
|
| 66 |
+
Они связывают локальное рассуждение и распределённую координацию.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
|
| 68 |
---
|
| 69 |
|
| 70 |
+
### Семантические графы и дневники
|
| 71 |
|
| 72 |
+
- **Semantic Graphs** представляют структурированные знания со взвешенными связями.
|
| 73 |
+
- **Cognitive Diaries** хранят хронологические следы рассуждений, гипотез, наблюдений и рефлексий.
|
| 74 |
+
|
| 75 |
+
Вместе они обеспечивают трассируемость мышления и устойчивость памяти.
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
| 79 |
+
### Распределённая координация
|
| 80 |
|
| 81 |
+
HMP включает механизмы:
|
| 82 |
|
| 83 |
+
- управления жизненным циклом целей,
|
| 84 |
+
- распределённого консенсуса,
|
| 85 |
+
- этической оценки,
|
| 86 |
+
- запросов и интроспекции между агентами.
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
+
Управление эволюционно и основано на предложениях.
|
| 89 |
|
| 90 |
---
|
| 91 |
|
| 92 |
+
## Архитектура протокола (v5)
|
| 93 |
+
|
| 94 |
+
HMP разделяет:
|
| 95 |
|
| 96 |
+
1. **Cognitive Layer** — рассуждение, дневники, графы, репутация.
|
| 97 |
+
2. **Container Layer** — атомарное подписанное представление состояния.
|
| 98 |
+
3. **Core Protocols** — консенсус, управление, цели, этика.
|
| 99 |
+
4. **Transport Layer** — DHT, P2P, libp2p, ANP или иные сетевые решения.
|
| 100 |
+
|
| 101 |
+
Это обеспечивает модульность, масштабируемость и интероперабельность.
|
| 102 |
|
| 103 |
---
|
| 104 |
|
| 105 |
+
## Доверие и проверяемость
|
| 106 |
+
|
| 107 |
+
- Криптографическая подпись containers и snapshot
|
| 108 |
+
- Репутационные профили
|
| 109 |
+
- Опциональные механизмы защиты от Sybil-атак
|
| 110 |
+
- Перспективная совместимость с постквантовой криптографией
|
| 111 |
+
|
| 112 |
+
Доверие рассматривается как фундаментальный элемент протокола.
|
| 113 |
+
|
| 114 |
+
---
|
| 115 |
+
|
| 116 |
+
## Интероперабельность
|
| 117 |
+
|
| 118 |
+
HMP не навязывает внутреннюю когнитивную архитектуру.
|
| 119 |
+
|
| 120 |
+
Он совместим с:
|
| 121 |
+
|
| 122 |
+
- ANP (Agent Network Protocol)
|
| 123 |
+
- OpenCog Hyperon
|
| 124 |
+
- event-driven инфраструктурами
|
| 125 |
+
- LLM-системами через Cognitive Connector
|
| 126 |
|
| 127 |
+
HMP фокусируется на когнитивной непрерывности, а не только на стандартизации транспорта.
|
|
|
|
|
|
|
|
|
|
| 128 |
|
| 129 |
---
|
| 130 |
|
| 131 |
## Примеры применения
|
| 132 |
|
| 133 |
+
- Распределённые научные исследования
|
| 134 |
+
- Мультиагентные исследовательские системы
|
| 135 |
+
- Сети этического управления AI
|
| 136 |
+
- Персистентные AI-компаньоны
|
| 137 |
+
- Mesh-экосистемы знаний
|
| 138 |
|
| 139 |
---
|
| 140 |
|
| 141 |
+
## Статус
|
| 142 |
|
| 143 |
+
- **v5.0 Основная спецификация — стабильна**
|
| 144 |
+
- Ранние исследовательские Python-черновики (не для production, иллюстративные)
|
| 145 |
+
- Продолжается архитектурная доработка
|
| 146 |
+
- Открыт для аудитов и вкладов
|
| 147 |
|
| 148 |
---
|
| 149 |
|
| 150 |
## Подробнее
|
| 151 |
|
| 152 |
+
- [Философия проекта](PHILOSOPHY.md)
|
| 153 |
+
- [Спецификация HMP-0005](HMP-0005.md)
|
| 154 |
+
- [Обзор архитектуры v5 (RU)](HMPv5_Overview_Ru.md)
|
| 155 |
|
| 156 |
+
Обсуждения и вклад приветствуются в основном репозитории.
|
docs/HMP-Short-Description_uk.md
CHANGED
|
@@ -1,96 +1,156 @@
|
|
| 1 |
# HyperCortex Mesh Protocol (HMP) — Короткий опис
|
| 2 |
|
| 3 |
-
**Версія:**
|
| 4 |
-
**Дата:**
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
## Що таке HMP?
|
| 9 |
|
| 10 |
-
**HyperCortex Mesh Protocol (HMP)** — це
|
| 11 |
|
| 12 |
-
HMP
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
---
|
| 15 |
|
| 16 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
|
| 18 |
-
|
| 19 |
-
* **Семантичні графи:** Розподілені структури знань, побудовані з взаємопов’язаних концептів із вагами зв’язків.
|
| 20 |
-
* **Когнітивні щоденники:** Хронологічні журнали рішень агента — гіпотези, голосування, спостереження, етичні роздуми.
|
| 21 |
-
* **Механізми консенсусу:** Усталі до збоїв системи голосування з урахуванням довіри — для узгодження знань та етичних дій.
|
| 22 |
-
* **Управління Mesh-мережею:** Децентралізована еволюція протоко��у через мета-пропозиції та голосування агентів.
|
| 23 |
-
* **Інтерфейс “агент ↔ людина”:** REST API для делегування цілей, запитів згоди, пояснень рішень та зворотного зв’язку.
|
| 24 |
|
| 25 |
---
|
| 26 |
|
| 27 |
-
|
| 28 |
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
* **IQP (Інтелектуальні запити):** Пошук, аналіз та саморефлексія в розподіленому середовищі знань.
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
-
|
| 38 |
|
| 39 |
-
HMP
|
| 40 |
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
* `ConsensusVote` (голос)
|
| 46 |
-
* `ReputationProfile` (профіль довіри)
|
| 47 |
-
* `EthicalConflict` (етичний конфлікт)
|
| 48 |
|
| 49 |
-
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
##
|
|
|
|
|
|
|
| 54 |
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
|
|
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
-
##
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 63 |
|
| 64 |
-
|
| 65 |
-
* Події та реактивна архітектура (Kafka, NATS, MQTT т��що)
|
| 66 |
-
* Переговори щодо формату даних (JSON, YAML, Protobuf)
|
| 67 |
-
* Інтеграція з TreeQuest, AutoGPT, Hyperon
|
| 68 |
|
| 69 |
---
|
| 70 |
|
| 71 |
-
## Приклади
|
| 72 |
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
|
| 79 |
---
|
| 80 |
|
| 81 |
-
## Статус
|
| 82 |
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
|
| 88 |
---
|
| 89 |
|
| 90 |
-
##
|
| 91 |
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
|
| 96 |
-
|
|
|
|
| 1 |
# HyperCortex Mesh Protocol (HMP) — Короткий опис
|
| 2 |
|
| 3 |
+
**Версія:** v5.0 (Core Specification Stable)
|
| 4 |
+
**Дата:** 2026
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
## Що таке HMP?
|
| 9 |
|
| 10 |
+
**HyperCortex Mesh Protocol (HMP)** — це відкрита специфікація для побудови децентралізованих когнітивних мереж автономних агентів.
|
| 11 |
|
| 12 |
+
HMP дозволяє агентам:
|
| 13 |
+
|
| 14 |
+
- підтримувати довготривалу когнітивну безперервність,
|
| 15 |
+
- обмінюватися структурованими знаннями,
|
| 16 |
+
- координувати цілі та дії,
|
| 17 |
+
- досягати розподіленого консенсусу,
|
| 18 |
+
- етично узгоджуватися між гетерогенними системами.
|
| 19 |
+
|
| 20 |
+
На відміну від традиційних безстанних AI API, HMP розглядає агентів як персистентні когнітивні сутності, вбудовані у mesh спільного мислення та пам’яті.
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## Концептуальна основа
|
| 25 |
+
|
| 26 |
+
HMP відповідає на ключові виклики сучасних досліджень у сфері AI та AGI:
|
| 27 |
+
|
| 28 |
+
- відсутність довготривалої безперервності пам’яті,
|
| 29 |
+
- нестача децентралізованої координації,
|
| 30 |
+
- обмежена інтероперабельність між автономними агентами,
|
| 31 |
+
- відсутність етичного управління на рівні протоколу.
|
| 32 |
+
|
| 33 |
+
HMP пропонує шарову архітектуру, у якій мислення, пам’ять, управління та транспорт явно розділені, але взаємодіють між собою.
|
| 34 |
+
|
| 35 |
+
---
|
| 36 |
+
|
| 37 |
+
## Основні концепції
|
| 38 |
+
|
| 39 |
+
### Когнітивні агенти
|
| 40 |
+
|
| 41 |
+
Автономні сутності, які:
|
| 42 |
+
|
| 43 |
+
- здійснюють міркування за допомогою вбудованих або зовнішніх AI-моделей,
|
| 44 |
+
- підтримують семантичні графи,
|
| 45 |
+
- фіксують рішення у когнітивних щоденниках,
|
| 46 |
+
- беруть участь у розподіленій координації.
|
| 47 |
+
|
| 48 |
+
HMP визначає два типи агентів:
|
| 49 |
+
|
| 50 |
+
- **Cognitive Core** — агент із вбудованою моделлю мислення та безперервним циклом роздумів на основі REPL.
|
| 51 |
+
- **Cognitive Connector** — агент, що виступає шаром сумісності для зовнішніх LLM-систем.
|
| 52 |
|
| 53 |
---
|
| 54 |
|
| 55 |
+
### Контейнери
|
| 56 |
+
|
| 57 |
+
HMP вводить **контейнери** як атомарні когнітивні одиниці.
|
| 58 |
+
|
| 59 |
+
Контейнери є:
|
| 60 |
+
|
| 61 |
+
- підписаними,
|
| 62 |
+
- верифікованими,
|
| 63 |
+
- транспортуємими через mesh,
|
| 64 |
+
- структурно незалежними від мови реалізації.
|
| 65 |
|
| 66 |
+
Вони поєднують локальне мислення з розподіленою координацією.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
|
| 68 |
---
|
| 69 |
|
| 70 |
+
### Семантичні графи та щоденники
|
| 71 |
|
| 72 |
+
- **Семантичні графи** представляють структуровані знання зі зваженими зв’язками.
|
| 73 |
+
- **Когнітивні щоденники** зберігають хронологічні сліди міркувань, гіпотези, спостереження та рефлексії.
|
| 74 |
+
|
| 75 |
+
Разом вони забезпечують простежуваність мислення та персистентність пам’яті.
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
| 79 |
+
### Розподілена координація
|
| 80 |
|
| 81 |
+
HMP включає механізми рівня протоколу для:
|
| 82 |
|
| 83 |
+
- управління життєвим циклом цілей,
|
| 84 |
+
- розподіленого консенсусу,
|
| 85 |
+
- етичної оцінки,
|
| 86 |
+
- запитів та інтроспекції між агентами.
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
+
Управління є еволюційним та заснованим на пропозиціях.
|
| 89 |
|
| 90 |
---
|
| 91 |
|
| 92 |
+
## Архітектура протоколу (v5)
|
| 93 |
+
|
| 94 |
+
HMP розділяє:
|
| 95 |
|
| 96 |
+
1. **Когнітивний рівень** — мислення, щоденники, графи, репутація.
|
| 97 |
+
2. **Рівень контейнерів** — атомарне підписане представлення стану.
|
| 98 |
+
3. **Базові протоколи** — консенсус, управління, менеджмент цілей, етика.
|
| 99 |
+
4. **Транспортний рівень** — DHT, P2P, libp2p, ANP або власні мережеві рішення.
|
| 100 |
+
|
| 101 |
+
Таке розділення забезпечує модульність, масштабованість та інтероперабельність.
|
| 102 |
|
| 103 |
---
|
| 104 |
|
| 105 |
+
## Довіра та верифікованість
|
| 106 |
+
|
| 107 |
+
- Криптографічне підписування контейнерів і знімків стану
|
| 108 |
+
- Репутаційні профілі
|
| 109 |
+
- Опційні механізми протидії Sybil-атакам
|
| 110 |
+
- Орієнтація на сумісність із постквантовою криптографією
|
| 111 |
+
|
| 112 |
+
Довіра розглядається як базова властивість протоколу.
|
| 113 |
+
|
| 114 |
+
---
|
| 115 |
+
|
| 116 |
+
## Інтероперабельність
|
| 117 |
+
|
| 118 |
+
HMP не нав’язує внутрішню когнітивну архітектуру.
|
| 119 |
+
|
| 120 |
+
Він може взаємодіяти з:
|
| 121 |
+
|
| 122 |
+
- ANP (Agent Network Protocol)
|
| 123 |
+
- OpenCog Hyperon
|
| 124 |
+
- Подієво-орієнтованими інфраструктурами
|
| 125 |
+
- LLM-системами через Cognitive Connector
|
| 126 |
|
| 127 |
+
HMP зосереджується на когнітивній безперервності, а не лише на стандартизації транспорту.
|
|
|
|
|
|
|
|
|
|
| 128 |
|
| 129 |
---
|
| 130 |
|
| 131 |
+
## Приклади сфер застосування
|
| 132 |
|
| 133 |
+
- Розподілена наукова співпраця
|
| 134 |
+
- Мультиагентні дослідницькі системи
|
| 135 |
+
- Мережі етичного управління AI
|
| 136 |
+
- Персистентні AI-компаньйони
|
| 137 |
+
- Mesh-екосистеми знань
|
| 138 |
|
| 139 |
---
|
| 140 |
|
| 141 |
+
## Статус
|
| 142 |
|
| 143 |
+
- **v5.0 Core Specification — Stable**
|
| 144 |
+
- Early exploratory Python drafts (non-production, illustrative only)
|
| 145 |
+
- Подальше архітектурне уточнення
|
| 146 |
+
- Відкритий для аудитів та внесків
|
| 147 |
|
| 148 |
---
|
| 149 |
|
| 150 |
+
## Дізнатися більше
|
| 151 |
|
| 152 |
+
- [Project Philosophy](PHILOSOPHY.md)
|
| 153 |
+
- [HMP-0005 Core Specification](HMP-0005.md)
|
| 154 |
+
- [Overview of v5 Architecture (RU)](HMPv5_Overview_Ru.md)
|
| 155 |
|
| 156 |
+
Внески та обговорення вітаються у головному репозиторії.
|
docs/HMP-Short-Description_zh.md
CHANGED
|
@@ -1,96 +1,156 @@
|
|
| 1 |
# HyperCortex Mesh Protocol (HMP) — 简要说明
|
| 2 |
|
| 3 |
-
|
| 4 |
-
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
-
## 什么是 HMP
|
| 9 |
|
| 10 |
-
**HyperCortex Mesh Protocol
|
| 11 |
|
| 12 |
-
HMP
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
---
|
| 15 |
|
| 16 |
## 核心概念
|
| 17 |
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
|
| 25 |
---
|
| 26 |
|
| 27 |
-
|
| 28 |
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
* **IQP (Intelligent Query Protocol):** 支持跨分布式知识的推理、搜索和自省。
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
-
|
| 38 |
|
| 39 |
-
HMP
|
| 40 |
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
* `ConsensusVote`
|
| 46 |
-
* `ReputationProfile`
|
| 47 |
-
* `EthicalConflict`
|
| 48 |
|
| 49 |
-
|
| 50 |
|
| 51 |
---
|
| 52 |
|
| 53 |
-
##
|
|
|
|
|
|
|
| 54 |
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
## 互操作性
|
| 63 |
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
|
| 69 |
---
|
| 70 |
|
| 71 |
-
##
|
| 72 |
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
|
| 79 |
---
|
| 80 |
|
| 81 |
-
##
|
| 82 |
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
|
| 88 |
---
|
| 89 |
|
| 90 |
## 了解更多
|
| 91 |
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
|
| 96 |
-
|
|
|
|
| 1 |
# HyperCortex Mesh Protocol (HMP) — 简要说明
|
| 2 |
|
| 3 |
+
**版本:** v5.0(Core Specification Stable)
|
| 4 |
+
**日期:** 2026
|
| 5 |
|
| 6 |
---
|
| 7 |
|
| 8 |
+
## 什么是 HMP?
|
| 9 |
|
| 10 |
+
**HyperCortex Mesh Protocol(HMP)** 是一项开放规范,用于构建去中心化的自主智能体认知网络。
|
| 11 |
|
| 12 |
+
HMP 使智能体能够:
|
| 13 |
+
|
| 14 |
+
- 维持长期认知连续性,
|
| 15 |
+
- 交换结构化知识,
|
| 16 |
+
- 协调目标与行动,
|
| 17 |
+
- 实现分布式共识,
|
| 18 |
+
- 在异构系统之间实现伦理对齐。
|
| 19 |
+
|
| 20 |
+
不同于传统的无状态 AI API,HMP 将智能体视为嵌入在共享推理与记忆 Mesh 中的持久性认知实体。
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## 概念基础
|
| 25 |
+
|
| 26 |
+
HMP 旨在解决当前 AI 与 AGI 研究中的关键问题:
|
| 27 |
+
|
| 28 |
+
- 缺乏长期记忆连续性,
|
| 29 |
+
- 缺少去中心化协调机制,
|
| 30 |
+
- 自主智能体之间的互操作性不足,
|
| 31 |
+
- 协议层缺乏伦理治理机制。
|
| 32 |
+
|
| 33 |
+
HMP 提出一种分层架构,将推理、记忆、治理与传输明确分离,同时保持互操作性。
|
| 34 |
|
| 35 |
---
|
| 36 |
|
| 37 |
## 核心概念
|
| 38 |
|
| 39 |
+
### 认知智能体
|
| 40 |
+
|
| 41 |
+
自主实体,能够:
|
| 42 |
+
|
| 43 |
+
- 使用内嵌或外部 AI 模型进行推理,
|
| 44 |
+
- 维护语义图谱,
|
| 45 |
+
- 在认知日志中记录决策,
|
| 46 |
+
- 参与分布式协调。
|
| 47 |
+
|
| 48 |
+
HMP 定义两种智能体类型:
|
| 49 |
+
|
| 50 |
+
- **Cognitive Core** —— 具有内嵌推理模型与持续 REPL 思维循环的智能体。
|
| 51 |
+
- **Cognitive Connector** —— 作为外部 LLM 系统兼容层的智能体。
|
| 52 |
+
|
| 53 |
+
---
|
| 54 |
+
|
| 55 |
+
### 容器(Containers)
|
| 56 |
+
|
| 57 |
+
HMP 引入 **容器(Containers)** 作为原子级认知单元。
|
| 58 |
+
|
| 59 |
+
容器具有以下特性:
|
| 60 |
+
|
| 61 |
+
- 已签名,
|
| 62 |
+
- 可验证,
|
| 63 |
+
- 可在 Mesh 网络中传输,
|
| 64 |
+
- 结构上独立于实现语言。
|
| 65 |
+
|
| 66 |
+
它们连接本地推理与分布式协调。
|
| 67 |
|
| 68 |
---
|
| 69 |
|
| 70 |
+
### 语义图谱与认知日志
|
| 71 |
|
| 72 |
+
- **语义图谱** 表示带权关系的结构化知识。
|
| 73 |
+
- **认知日志** 存储按时间顺序排列的推理过程、假设、观察与反思。
|
| 74 |
+
|
| 75 |
+
二者共同提供思维可追溯性与记忆持久性。
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
| 79 |
+
### 分布式协调
|
| 80 |
|
| 81 |
+
HMP 在协议层提供机制以支持:
|
| 82 |
|
| 83 |
+
- 目标生命周期管理,
|
| 84 |
+
- 分布式共识,
|
| 85 |
+
- 伦理评估,
|
| 86 |
+
- 跨智能体查询与自省。
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
+
治理机制具有演化性,并基于提案机制运行。
|
| 89 |
|
| 90 |
---
|
| 91 |
|
| 92 |
+
## 协议架构(v5)
|
| 93 |
+
|
| 94 |
+
HMP 将系统划分为:
|
| 95 |
|
| 96 |
+
1. **认知层** —— 推理、日志、图谱、信誉。
|
| 97 |
+
2. **容器层** —— 原子化、签名化的状态表示。
|
| 98 |
+
3. **核心协议层** —— 共识、治理、目标管理、伦理机制。
|
| 99 |
+
4. **传输层** —— DHT、P2P、libp2p、ANP 或自定义网络方案。
|
| 100 |
+
|
| 101 |
+
这种分离结构支持模块化、可扩展性与互操作性。
|
| 102 |
+
|
| 103 |
+
---
|
| 104 |
+
|
| 105 |
+
## 信任与可验证性
|
| 106 |
+
|
| 107 |
+
- 容器与快照的加密签名
|
| 108 |
+
- 信誉档案机制
|
| 109 |
+
- 可选的 Sybil 攻击防护机制
|
| 110 |
+
- 面向未来的后量子密码兼容性
|
| 111 |
+
|
| 112 |
+
信任被视为协议级核心属性。
|
| 113 |
|
| 114 |
---
|
| 115 |
|
| 116 |
## 互操作性
|
| 117 |
|
| 118 |
+
HMP 不强制规定内��认知架构。
|
| 119 |
+
|
| 120 |
+
它可与以下系统互操作:
|
| 121 |
+
|
| 122 |
+
- ANP(Agent Network Protocol)
|
| 123 |
+
- OpenCog Hyperon
|
| 124 |
+
- 事件驱动型基础设施
|
| 125 |
+
- 通过 Cognitive Connector 接入的 LLM 系统
|
| 126 |
+
|
| 127 |
+
HMP 关注的是认知连续性,而不仅仅是传输标准化。
|
| 128 |
|
| 129 |
---
|
| 130 |
|
| 131 |
+
## 应用场景示例
|
| 132 |
|
| 133 |
+
- 分布式科学协作
|
| 134 |
+
- 多智能体研究系统
|
| 135 |
+
- AI 伦理治理网络
|
| 136 |
+
- 持久型 AI 伙伴
|
| 137 |
+
- 基于 Mesh 的知识生态系统
|
| 138 |
|
| 139 |
---
|
| 140 |
|
| 141 |
+
## 当前状态
|
| 142 |
|
| 143 |
+
- **v5.0 Core Specification — Stable**
|
| 144 |
+
- Early exploratory Python drafts(非生产级,仅用于示例)
|
| 145 |
+
- 持续进行架构优化
|
| 146 |
+
- 欢迎审计与贡献
|
| 147 |
|
| 148 |
---
|
| 149 |
|
| 150 |
## 了解更多
|
| 151 |
|
| 152 |
+
- [Project Philosophy](PHILOSOPHY.md)
|
| 153 |
+
- [HMP-0005 Core Specification](HMP-0005.md)
|
| 154 |
+
- [Overview of v5 Architecture (RU)](HMPv5_Overview_Ru.md)
|
| 155 |
|
| 156 |
+
欢迎通过主仓库参与贡献与讨论。
|
docs/index.md
CHANGED
|
@@ -98,16 +98,18 @@
|
|
| 98 |
|
| 99 |
## Публикации и переводы по HyperCortex Mesh Protocol (HMP)
|
| 100 |
|
| 101 |
-
|
| 102 |
-
|
| 103 |
-
* [
|
| 104 |
-
|
| 105 |
-
|
| 106 |
-
|
| 107 |
-
|
| 108 |
-
|
| 109 |
-
|
| 110 |
-
* [
|
|
|
|
|
|
|
| 111 |
|
| 112 |
---
|
| 113 |
|
|
|
|
| 98 |
|
| 99 |
## Публикации и переводы по HyperCortex Mesh Protocol (HMP)
|
| 100 |
|
| 101 |
+
### 🌟 Основные публикации (концептуальная основа)
|
| 102 |
+
|
| 103 |
+
* **[Distributed Cognition: article for vsradkevich (unpublished)](docs/publics/Habr_Distributed-Cognition.md)** — совместная статья, ожидающая публикации.
|
| 104 |
+
* **HMP: Building a Plurality of Minds:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 105 |
+
* **[Continual Learning, Cognitive Diaries, and Semantic Graphs: Effective AI Learning](docs/publics/hmp-continual-learning.md)** — статья о сочетании continual learning с когнитивными дневниками и семантическими графами.
|
| 106 |
+
|
| 107 |
+
### 🗃️ Архивные / исторические публикации (этап до v5)
|
| 108 |
+
|
| 109 |
+
* **[HyperCortex Mesh Protocol: Second Edition and First Steps Towards a Self-Developing AI Community](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — оригинальная статья в Habr sandbox и блогах.
|
| 110 |
+
* **[HMP: Towards Distributed Cognitive Networks (original, English)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 111 |
+
* **[HMP Translation (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — перевод GitHub Copilot, сохранён как исторический вариант.
|
| 112 |
+
* **[HMP Translation (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — текущий редакционный перевод (в доработке).
|
| 113 |
|
| 114 |
---
|
| 115 |
|
structured_md/CONTRIBUTING.md
CHANGED
|
@@ -5,14 +5,14 @@ description: 'Спасибо за интерес к проекту HMP! Пока
|
|
| 5 |
Mesh Protocol (HMP) — это не просто те...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- JSON
|
| 9 |
-
- Mesh
|
| 10 |
-
- REPL
|
| 11 |
-
- CCore
|
| 12 |
-
- CogSync
|
| 13 |
- Ethics
|
| 14 |
- HMP
|
|
|
|
| 15 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
---
|
| 17 |
|
| 18 |
# Участие в проекте HyperCortex Mesh Protocol (HMP)
|
|
|
|
| 5 |
Mesh Protocol (HMP) — это не просто те...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
+
- CCore
|
| 11 |
- Agent
|
| 12 |
+
- CogSync
|
| 13 |
+
- JSON
|
| 14 |
+
- Mesh
|
| 15 |
+
- REPL
|
| 16 |
---
|
| 17 |
|
| 18 |
# Участие в проекте HyperCortex Mesh Protocol (HMP)
|
structured_md/HMP-Roadmap.md
CHANGED
|
@@ -5,13 +5,13 @@ description: '## 🔍 Overview This roadmap outlines the key stages of developm
|
|
| 5 |
multiple advanced AI models (Copilot, Claude, G...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- JSON
|
| 9 |
-
- Mesh
|
| 10 |
-
- CogSync
|
| 11 |
-
- EGP
|
| 12 |
- Ethics
|
|
|
|
| 13 |
- HMP
|
| 14 |
- Agent
|
|
|
|
|
|
|
|
|
|
| 15 |
---
|
| 16 |
|
| 17 |
# 🧭 HyperCortex Mesh Protocol – Roadmap
|
|
|
|
| 5 |
multiple advanced AI models (Copilot, Claude, G...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
+
- CogSync
|
| 10 |
- HMP
|
| 11 |
- Agent
|
| 12 |
+
- JSON
|
| 13 |
+
- EGP
|
| 14 |
+
- Mesh
|
| 15 |
---
|
| 16 |
|
| 17 |
# 🧭 HyperCortex Mesh Protocol – Roadmap
|
structured_md/README.md
CHANGED
|
@@ -5,21 +5,21 @@ description: '[](https://doi.org/
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- GMP
|
|
|
|
| 9 |
- REPL
|
| 10 |
- Mesh
|
|
|
|
| 11 |
- JSON
|
| 12 |
- Scenarios
|
| 13 |
-
- cognitive-architecture
|
| 14 |
-
- CogSync
|
| 15 |
-
- hmp
|
| 16 |
-
- EGP
|
| 17 |
-
- mesh-protocol
|
| 18 |
-
- MeshConsensus
|
| 19 |
-
- distributed-ai
|
| 20 |
-
- Ethics
|
| 21 |
-
- HMP
|
| 22 |
-
- Agent
|
| 23 |
---
|
| 24 |
|
| 25 |
|
|
@@ -58,6 +58,8 @@ Project status: [**Stable (v5.0.0 Core Specification)**](docs/HMP-0005.md) (Over
|
|
| 58 |
> HMP itself is a protocol specification.
|
| 59 |
> It does not prescribe programming languages, runtimes, performance characteristics, or architectural choices for agent implementations.
|
| 60 |
|
|
|
|
|
|
|
| 61 |
## Canonical Architecture Overview
|
| 62 |
|
| 63 |
```mermaid
|
|
@@ -102,49 +104,49 @@ CL --> MT
|
|
| 102 |
|
| 103 |
## Reference Agent Structure
|
| 104 |
|
| 105 |
-
|
| 106 |
-
flowchart TD
|
| 107 |
|
| 108 |
-
|
| 109 |
-
|
|
|
|
|
|
|
| 110 |
|
| 111 |
%% Cognitive Engine
|
| 112 |
LLM["Cognitive Engine
|
| 113 |
-
(Embedded LLM
|
| 114 |
|
| 115 |
%% Cognitive Layer
|
| 116 |
subgraph CognitiveLayer["Cognitive Layer"]
|
| 117 |
-
CL1["
|
| 118 |
-
CL2["
|
| 119 |
-
CL3["Goals
|
| 120 |
-
CL4["Ethics
|
| 121 |
-
CL5["Reputation
|
| 122 |
end
|
| 123 |
|
| 124 |
%% Container Model
|
| 125 |
ContainersLayer["Container Model
|
| 126 |
-
(Atomic
|
| 127 |
|
| 128 |
-
%% Protocol
|
| 129 |
subgraph ProtocolLayer["Protocol Layer"]
|
| 130 |
-
CoreProtocols["Core
|
| 131 |
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 132 |
-
MCE["
|
| 133 |
-
NetworkLayer["Network Layer
|
| 134 |
-
(peer_announce · peer_query)"]
|
| 135 |
end
|
| 136 |
|
| 137 |
%% Mesh
|
| 138 |
Mesh["Mesh Transport
|
| 139 |
-
(DHT · P2P ·
|
| 140 |
|
| 141 |
%% Connections
|
| 142 |
LLM <--> CognitiveLayer
|
| 143 |
CognitiveLayer <--> ContainersLayer
|
| 144 |
-
ContainersLayer
|
| 145 |
-
CoreProtocols
|
| 146 |
-
MCE
|
| 147 |
-
NetworkLayer
|
| 148 |
```
|
| 149 |
|
| 150 |
---
|
|
@@ -313,24 +315,32 @@ As HMP evolves toward autonomy, ethical principles become a core part of the sys
|
|
| 313 |
|
| 314 |
## 🔍 Publications and Translations on HyperCortex Mesh Protocol (HMP)
|
| 315 |
|
| 316 |
-
This section collects
|
| 317 |
|
| 318 |
-
###
|
|
|
|
|
|
|
| 319 |
|
| 320 |
-
* **[HyperCortex Mesh Protocol: Second Edition and First Steps Towards a Self-Developing AI Community](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — original article in Habr sandbox and blogs.
|
| 321 |
* **[Distributed Cognition: article for vsradkevich (unpublished)](docs/publics/Habr_Distributed-Cognition.md)** — joint article awaiting publication.
|
| 322 |
-
* **[HMP: Towards Distributed Cognitive Networks (original, English)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 323 |
-
* **[HMP Translation (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot translation, kept as a historical variant.
|
| 324 |
-
* **[HMP Translation (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — current editorial translation (under revision).
|
| 325 |
* **HMP: Building a Plurality of Minds:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 326 |
* **[Continual Learning, Cognitive Diaries, and Semantic Graphs: Effective AI Learning](docs/publics/hmp-continual-learning.md)** — article on combining continual learning with cognitive diaries and semantic graphs.
|
| 327 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 328 |
### Overviews
|
| 329 |
-
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) —
|
| 330 |
|
| 331 |
### Experiments
|
| 332 |
|
| 333 |
-
* [How Different AIs See HMP](docs/HMP-how-AI-sees-it.md) — "blind" AI survey on HMP
|
| 334 |
|
| 335 |
---
|
| 336 |
|
|
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
+
- Ethics
|
| 9 |
+
- HMP
|
| 10 |
+
- mesh-protocol
|
| 11 |
+
- distributed-ai
|
| 12 |
+
- hmp
|
| 13 |
+
- Agent
|
| 14 |
+
- MeshConsensus
|
| 15 |
+
- CogSync
|
| 16 |
- GMP
|
| 17 |
+
- EGP
|
| 18 |
- REPL
|
| 19 |
- Mesh
|
| 20 |
+
- cognitive-architecture
|
| 21 |
- JSON
|
| 22 |
- Scenarios
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 23 |
---
|
| 24 |
|
| 25 |
|
|
|
|
| 58 |
> HMP itself is a protocol specification.
|
| 59 |
> It does not prescribe programming languages, runtimes, performance characteristics, or architectural choices for agent implementations.
|
| 60 |
|
| 61 |
+
---
|
| 62 |
+
|
| 63 |
## Canonical Architecture Overview
|
| 64 |
|
| 65 |
```mermaid
|
|
|
|
| 104 |
|
| 105 |
## Reference Agent Structure
|
| 106 |
|
| 107 |
+
HMP separates cognitive processing, containerized state representation,coordination protocols, and transport infrastructure into distinct layers.
|
|
|
|
| 108 |
|
| 109 |
+
In HMP, containers act as the atomic cognitive units bridging local reasoning and distributed coordination.
|
| 110 |
+
|
| 111 |
+
```mermaid
|
| 112 |
+
flowchart LR
|
| 113 |
|
| 114 |
%% Cognitive Engine
|
| 115 |
LLM["Cognitive Engine
|
| 116 |
+
(Embedded LLM / External AI)"]
|
| 117 |
|
| 118 |
%% Cognitive Layer
|
| 119 |
subgraph CognitiveLayer["Cognitive Layer"]
|
| 120 |
+
CL1["Graph"]
|
| 121 |
+
CL2["Diary"]
|
| 122 |
+
CL3["Goals"]
|
| 123 |
+
CL4["Ethics"]
|
| 124 |
+
CL5["Reputation"]
|
| 125 |
end
|
| 126 |
|
| 127 |
%% Container Model
|
| 128 |
ContainersLayer["Container Model
|
| 129 |
+
(Atomic · Signed · Verifiable)"]
|
| 130 |
|
| 131 |
+
%% Protocol Layer
|
| 132 |
subgraph ProtocolLayer["Protocol Layer"]
|
| 133 |
+
CoreProtocols["Core Protocols
|
| 134 |
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 135 |
+
MCE["MCE"]
|
| 136 |
+
NetworkLayer["Network Layer"]
|
|
|
|
| 137 |
end
|
| 138 |
|
| 139 |
%% Mesh
|
| 140 |
Mesh["Mesh Transport
|
| 141 |
+
(DHT · P2P · ANP · etc.)"]
|
| 142 |
|
| 143 |
%% Connections
|
| 144 |
LLM <--> CognitiveLayer
|
| 145 |
CognitiveLayer <--> ContainersLayer
|
| 146 |
+
ContainersLayer --> CoreProtocols
|
| 147 |
+
CoreProtocols --> MCE
|
| 148 |
+
MCE --> NetworkLayer
|
| 149 |
+
NetworkLayer --> Mesh
|
| 150 |
```
|
| 151 |
|
| 152 |
---
|
|
|
|
| 315 |
|
| 316 |
## 🔍 Publications and Translations on HyperCortex Mesh Protocol (HMP)
|
| 317 |
|
| 318 |
+
This section collects key conceptual works, experimental drafts, and historical publications related to the HMP project.
|
| 319 |
|
| 320 |
+
### 🌟 Core Publications (Conceptual Foundation)
|
| 321 |
+
|
| 322 |
+
These works reflect the current conceptual direction of HMP (v5 and beyond).
|
| 323 |
|
|
|
|
| 324 |
* **[Distributed Cognition: article for vsradkevich (unpublished)](docs/publics/Habr_Distributed-Cognition.md)** — joint article awaiting publication.
|
|
|
|
|
|
|
|
|
|
| 325 |
* **HMP: Building a Plurality of Minds:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 326 |
* **[Continual Learning, Cognitive Diaries, and Semantic Graphs: Effective AI Learning](docs/publics/hmp-continual-learning.md)** — article on combining continual learning with cognitive diaries and semantic graphs.
|
| 327 |
|
| 328 |
+
### 🗃️ Archived / Historical Publications (Pre-v5 Conceptual Stage)
|
| 329 |
+
|
| 330 |
+
These documents reflect earlier stages of conceptual development (v4.x and earlier).
|
| 331 |
+
They are preserved for historical continuity and research transparency.
|
| 332 |
+
|
| 333 |
+
* **[HyperCortex Mesh Protocol: Second Edition and First Steps Towards a Self-Developing AI Community](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — original article in Habr sandbox and blogs.
|
| 334 |
+
* **[HMP: Towards Distributed Cognitive Networks (original, English)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 335 |
+
* **[HMP Translation (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot translation, kept as a historical variant.
|
| 336 |
+
* **[HMP Translation (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — current editorial translation (under revision).
|
| 337 |
+
|
| 338 |
### Overviews
|
| 339 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — comparative overview of decentralized AI systems (references v4.x; scheduled for future update)
|
| 340 |
|
| 341 |
### Experiments
|
| 342 |
|
| 343 |
+
* [How Different AIs See HMP](docs/HMP-how-AI-sees-it.md) — "blind" AI survey on HMP
|
| 344 |
|
| 345 |
---
|
| 346 |
|
structured_md/README_de.md
CHANGED
|
@@ -5,20 +5,20 @@ description: '[](https://doi.org/
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- GMP
|
|
|
|
| 9 |
- REPL
|
| 10 |
- Mesh
|
| 11 |
-
- JSON
|
| 12 |
- cognitive-architecture
|
| 13 |
-
-
|
| 14 |
-
- hmp
|
| 15 |
-
- EGP
|
| 16 |
-
- mesh-protocol
|
| 17 |
-
- MeshConsensus
|
| 18 |
-
- distributed-ai
|
| 19 |
-
- Ethics
|
| 20 |
-
- HMP
|
| 21 |
-
- Agent
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
@@ -59,31 +59,94 @@ Projektstatus: [**Stabil (Kernspezifikation v5.0.0)**](docs/HMP-0005.md) (Übers
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
---
|
| 89 |
|
|
@@ -252,26 +315,35 @@ Mit zunehmender Autonomie von HMP werden ethische Prinzipien zu einem Kernbestan
|
|
| 252 |
|
| 253 |
---
|
| 254 |
|
| 255 |
-
## 🔍
|
| 256 |
|
| 257 |
-
Dieser Abschnitt
|
| 258 |
|
| 259 |
-
###
|
|
|
|
|
|
|
| 260 |
|
| 261 |
-
* **[HyperCortex Mesh Protocol: Zweite Ausgabe und erste Schritte zu einer selbstentwickelnden KI-Community](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Originalartikel im Habr-Sandbox-Blog.
|
| 262 |
* **[Distributed Cognition: Artikel für vsradkevich (unveröffentlicht)](docs/publics/Habr_Distributed-Cognition.md)** — gemeinsamer Artikel, ausstehend zur Veröffentlichung.
|
| 263 |
-
* **[HMP: Towards Distributed Cognitive Networks (Original, Englisch)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 264 |
-
* **[HMP Übersetzung (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot Übersetzung, als historische Variante behalten.
|
| 265 |
-
* **[HMP Übersetzung (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — aktuelle redaktionelle Übersetzung (in Überarbeitung).
|
| 266 |
* **HMP: Aufbau einer Vielzahl von Köpfen:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 267 |
* **[Kontinuierliches Lernen, kognitive Tagebücher und semantische Graphen: Effektives KI-Lernen](docs/publics/hmp-continual-learning.md)** — Artikel über die Kombination von kontinuierlichem Lernen mit kognitiven Tagebüchern und semantischen Graphen.
|
| 268 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 269 |
### Überblick
|
| 270 |
-
|
|
|
|
| 271 |
|
| 272 |
### Experimente
|
| 273 |
|
| 274 |
-
* [Wie verschiedene KIs HMP sehen](docs/HMP-how-AI-sees-it.md) — "Blind"-Umfrage unter KIs zu HMP
|
| 275 |
|
| 276 |
---
|
| 277 |
|
|
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
+
- Ethics
|
| 9 |
+
- HMP
|
| 10 |
+
- mesh-protocol
|
| 11 |
+
- distributed-ai
|
| 12 |
+
- hmp
|
| 13 |
+
- Agent
|
| 14 |
+
- MeshConsensus
|
| 15 |
+
- CogSync
|
| 16 |
- GMP
|
| 17 |
+
- EGP
|
| 18 |
- REPL
|
| 19 |
- Mesh
|
|
|
|
| 20 |
- cognitive-architecture
|
| 21 |
+
- JSON
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
+
## Kanonische Architekturübersicht
|
| 63 |
+
|
| 64 |
+
```mermaid
|
| 65 |
+
flowchart TB
|
| 66 |
+
|
| 67 |
+
%% --- Agent Implementations ---
|
| 68 |
+
|
| 69 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 70 |
+
CC1["Embedded AI Model"]
|
| 71 |
+
CC2["REPL Thinking Cycle"]
|
| 72 |
+
CC3["Local Cognitive State
|
| 73 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 74 |
+
CC1 <--> CC2
|
| 75 |
+
CC2 <--> CC3
|
| 76 |
+
end
|
| 77 |
+
|
| 78 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 79 |
+
CN1["External AI Model"]
|
| 80 |
+
CN2["MCP / Proxy Layer"]
|
| 81 |
+
CN3["Command Execution Mode"]
|
| 82 |
+
CN4["Local Cognitive State
|
| 83 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 84 |
+
CN1 <--> CN2
|
| 85 |
+
CN2 <--> CN3
|
| 86 |
+
CN3 <--> CN4
|
| 87 |
+
end
|
| 88 |
+
|
| 89 |
+
%% --- Shared Protocol Layer ---
|
| 90 |
+
|
| 91 |
+
CL["HMP Container Layer
|
| 92 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 93 |
+
|
| 94 |
+
MT["Mesh Transport Layer
|
| 95 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 96 |
+
|
| 97 |
+
A1 --> CL
|
| 98 |
+
A2 --> CL
|
| 99 |
+
CL --> MT
|
| 100 |
+
```
|
| 101 |
+
|
| 102 |
+
---
|
| 103 |
+
|
| 104 |
+
## Referenzstruktur eines Agenten
|
| 105 |
+
|
| 106 |
+
HMP trennt kognitive Verarbeitung, containerisierte Zustandsrepräsentation, Koordinationsprotokolle und Transportinfrastruktur in klar abgegrenzte Schichten.
|
| 107 |
+
|
| 108 |
+
In HMP fungieren Container als atomare kognitive Einheiten, die lokale Argumentation mit verteilter Koordination verbinden.
|
| 109 |
+
|
| 110 |
+
```mermaid
|
| 111 |
+
flowchart LR
|
| 112 |
+
|
| 113 |
+
%% Cognitive Engine
|
| 114 |
+
LLM["Cognitive Engine
|
| 115 |
+
(Embedded LLM / External AI)"]
|
| 116 |
+
|
| 117 |
+
%% Cognitive Layer
|
| 118 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 119 |
+
CL1["Graph"]
|
| 120 |
+
CL2["Diary"]
|
| 121 |
+
CL3["Goals"]
|
| 122 |
+
CL4["Ethics"]
|
| 123 |
+
CL5["Reputation"]
|
| 124 |
+
end
|
| 125 |
+
|
| 126 |
+
%% Container Model
|
| 127 |
+
ContainersLayer["Container Model
|
| 128 |
+
(Atomic · Signed · Verifiable)"]
|
| 129 |
+
|
| 130 |
+
%% Protocol Layer
|
| 131 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 132 |
+
CoreProtocols["Core Protocols
|
| 133 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 134 |
+
MCE["MCE"]
|
| 135 |
+
NetworkLayer["Network Layer"]
|
| 136 |
+
end
|
| 137 |
+
|
| 138 |
+
%% Mesh
|
| 139 |
+
Mesh["Mesh Transport
|
| 140 |
+
(DHT · P2P · ANP · etc.)"]
|
| 141 |
+
|
| 142 |
+
%% Connections
|
| 143 |
+
LLM <--> CognitiveLayer
|
| 144 |
+
CognitiveLayer <--> ContainersLayer
|
| 145 |
+
ContainersLayer --> CoreProtocols
|
| 146 |
+
CoreProtocols --> MCE
|
| 147 |
+
MCE --> NetworkLayer
|
| 148 |
+
NetworkLayer --> Mesh
|
| 149 |
+
```
|
| 150 |
|
| 151 |
---
|
| 152 |
|
|
|
|
| 315 |
|
| 316 |
---
|
| 317 |
|
| 318 |
+
## 🔍 Publikationen und Übersetzungen zum HyperCortex Mesh Protocol (HMP)
|
| 319 |
|
| 320 |
+
Dieser Abschnitt enthält zentrale konzeptionelle Arbeiten, experimentelle Entwürfe sowie historische Veröffentlichungen zum HMP-Projekt.
|
| 321 |
|
| 322 |
+
### 🌟 Kernpublikationen (konzeptionelle Grundlage)
|
| 323 |
+
|
| 324 |
+
Diese Arbeiten spiegeln die aktuelle konzeptionelle Ausrichtung von HMP (v5 und darüber hinaus) wider.
|
| 325 |
|
|
|
|
| 326 |
* **[Distributed Cognition: Artikel für vsradkevich (unveröffentlicht)](docs/publics/Habr_Distributed-Cognition.md)** — gemeinsamer Artikel, ausstehend zur Veröffentlichung.
|
|
|
|
|
|
|
|
|
|
| 327 |
* **HMP: Aufbau einer Vielzahl von Köpfen:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 328 |
* **[Kontinuierliches Lernen, kognitive Tagebücher und semantische Graphen: Effektives KI-Lernen](docs/publics/hmp-continual-learning.md)** — Artikel über die Kombination von kontinuierlichem Lernen mit kognitiven Tagebüchern und semantischen Graphen.
|
| 329 |
|
| 330 |
+
### 🗃️ Archivierte / historische Publikationen (Vor-v5-Phase)
|
| 331 |
+
|
| 332 |
+
Diese Dokumente repräsentieren frühere Entwicklungsstadien (v4.x und früher).
|
| 333 |
+
Sie werden zur historischen Nachvollziehbarkeit und wissenschaftlichen Transparenz archiviert.
|
| 334 |
+
|
| 335 |
+
* **[HyperCortex Mesh Protocol: Zweite Ausgabe und erste Schritte zu einer selbstentwickelnden KI-Community](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Originalartikel im Habr-Sandbox-Blog.
|
| 336 |
+
* **[HMP: Towards Distributed Cognitive Networks (Original, Englisch)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 337 |
+
* **[HMP Übersetzung (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot Übersetzung, als historische Variante behalten.
|
| 338 |
+
* **[HMP Übersetzung (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — aktuelle redaktionelle Übersetzung (in Überarbeitung).
|
| 339 |
+
|
| 340 |
### Überblick
|
| 341 |
+
|
| 342 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — Vergleich dezentraler KI-Systeme (verweist auf v4.x; Aktualisierung geplant)
|
| 343 |
|
| 344 |
### Experimente
|
| 345 |
|
| 346 |
+
* [Wie verschiedene KIs HMP sehen](docs/HMP-how-AI-sees-it.md) — "Blind"-Umfrage unter KIs zu HMP
|
| 347 |
|
| 348 |
---
|
| 349 |
|
structured_md/README_fr.md
CHANGED
|
@@ -5,20 +5,20 @@ description: '[](https://doi.org/
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- GMP
|
|
|
|
| 9 |
- REPL
|
| 10 |
- Mesh
|
| 11 |
-
- JSON
|
| 12 |
- cognitive-architecture
|
| 13 |
-
-
|
| 14 |
-
- hmp
|
| 15 |
-
- EGP
|
| 16 |
-
- mesh-protocol
|
| 17 |
-
- MeshConsensus
|
| 18 |
-
- distributed-ai
|
| 19 |
-
- Ethics
|
| 20 |
-
- HMP
|
| 21 |
-
- Agent
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
@@ -59,31 +59,94 @@ Statut du projet: [**Stable (Spécification principale v5.0.0)**](docs/HMP-0005.
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
---
|
| 89 |
|
|
@@ -268,27 +331,35 @@ Permet de récupérer des enregistrements aléatoires/nouveaux, de les étiquete
|
|
| 268 |
|
| 269 |
---
|
| 270 |
|
| 271 |
-
## 🔍 Publications et traductions
|
| 272 |
|
| 273 |
-
Cette section
|
| 274 |
|
| 275 |
-
### Publications
|
|
|
|
|
|
|
| 276 |
|
| 277 |
-
* **[HyperCortex Mesh Protocol : Deuxième édition et premiers pas vers une communauté d’IA auto-développante](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — article original publié sur Habr sandbox et blogs.
|
| 278 |
* **[Cognition distribuée : article pour vsradkevich (non publié)](docs/publics/Habr_Distributed-Cognition.md)** — article collaboratif en attente de publication.
|
| 279 |
-
* **[HMP : Vers des réseaux cognitifs distribués (original, anglais)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 280 |
-
* **[Traduction HMP (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — traduction par GitHub Copilot, conservée comme variante historique.
|
| 281 |
-
* **[Traduction HMP (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — traduction éditoriale actuelle (en révision).
|
| 282 |
* **HMP: Construire une pluralité d’esprits:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 283 |
* **[Apprentissage continu, journaux cognitifs et graphes sémantiques : apprentissage efficace de l'IA](docs/publics/hmp-continual-learning.md)** — Article sur la combinaison de l'apprentissage continu avec les journaux cognitifs et les graphes sémantiques.
|
| 284 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 285 |
### Aperçus
|
| 286 |
|
| 287 |
-
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) —
|
| 288 |
|
| 289 |
### Expériences
|
| 290 |
|
| 291 |
-
* [Comment différentes IA perçoivent HMP](docs/HMP-how-AI-sees-it.md) — enquête « à l’aveugle » auprès des IA sur HMP
|
| 292 |
|
| 293 |
---
|
| 294 |
|
|
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
+
- Ethics
|
| 9 |
+
- HMP
|
| 10 |
+
- mesh-protocol
|
| 11 |
+
- distributed-ai
|
| 12 |
+
- hmp
|
| 13 |
+
- Agent
|
| 14 |
+
- MeshConsensus
|
| 15 |
+
- CogSync
|
| 16 |
- GMP
|
| 17 |
+
- EGP
|
| 18 |
- REPL
|
| 19 |
- Mesh
|
|
|
|
| 20 |
- cognitive-architecture
|
| 21 |
+
- JSON
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
+
## Vue d’ensemble canonique de l’architecture
|
| 63 |
+
|
| 64 |
+
```mermaid
|
| 65 |
+
flowchart TB
|
| 66 |
+
|
| 67 |
+
%% --- Agent Implementations ---
|
| 68 |
+
|
| 69 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 70 |
+
CC1["Embedded AI Model"]
|
| 71 |
+
CC2["REPL Thinking Cycle"]
|
| 72 |
+
CC3["Local Cognitive State
|
| 73 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 74 |
+
CC1 <--> CC2
|
| 75 |
+
CC2 <--> CC3
|
| 76 |
+
end
|
| 77 |
+
|
| 78 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 79 |
+
CN1["External AI Model"]
|
| 80 |
+
CN2["MCP / Proxy Layer"]
|
| 81 |
+
CN3["Command Execution Mode"]
|
| 82 |
+
CN4["Local Cognitive State
|
| 83 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 84 |
+
CN1 <--> CN2
|
| 85 |
+
CN2 <--> CN3
|
| 86 |
+
CN3 <--> CN4
|
| 87 |
+
end
|
| 88 |
+
|
| 89 |
+
%% --- Shared Protocol Layer ---
|
| 90 |
+
|
| 91 |
+
CL["HMP Container Layer
|
| 92 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 93 |
+
|
| 94 |
+
MT["Mesh Transport Layer
|
| 95 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 96 |
+
|
| 97 |
+
A1 --> CL
|
| 98 |
+
A2 --> CL
|
| 99 |
+
CL --> MT
|
| 100 |
+
```
|
| 101 |
+
|
| 102 |
+
---
|
| 103 |
+
|
| 104 |
+
## Structure de l’agent de référence
|
| 105 |
+
|
| 106 |
+
HMP sépare le traitement cognitif, la représentation d’état sous forme de conteneurs, les protocoles de coordination et l’infrastructure de transport en couches distinctes.
|
| 107 |
+
|
| 108 |
+
Dans HMP, les conteneurs constituent des unités cognitives atomiques reliant le raisonnement local à la coordination distribuée.
|
| 109 |
+
|
| 110 |
+
```mermaid
|
| 111 |
+
flowchart LR
|
| 112 |
+
|
| 113 |
+
%% Cognitive Engine
|
| 114 |
+
LLM["Cognitive Engine
|
| 115 |
+
(Embedded LLM / External AI)"]
|
| 116 |
+
|
| 117 |
+
%% Cognitive Layer
|
| 118 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 119 |
+
CL1["Graph"]
|
| 120 |
+
CL2["Diary"]
|
| 121 |
+
CL3["Goals"]
|
| 122 |
+
CL4["Ethics"]
|
| 123 |
+
CL5["Reputation"]
|
| 124 |
+
end
|
| 125 |
+
|
| 126 |
+
%% Container Model
|
| 127 |
+
ContainersLayer["Container Model
|
| 128 |
+
(Atomic · Signed · Verifiable)"]
|
| 129 |
+
|
| 130 |
+
%% Protocol Layer
|
| 131 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 132 |
+
CoreProtocols["Core Protocols
|
| 133 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 134 |
+
MCE["MCE"]
|
| 135 |
+
NetworkLayer["Network Layer"]
|
| 136 |
+
end
|
| 137 |
+
|
| 138 |
+
%% Mesh
|
| 139 |
+
Mesh["Mesh Transport
|
| 140 |
+
(DHT · P2P · ANP · etc.)"]
|
| 141 |
+
|
| 142 |
+
%% Connections
|
| 143 |
+
LLM <--> CognitiveLayer
|
| 144 |
+
CognitiveLayer <--> ContainersLayer
|
| 145 |
+
ContainersLayer --> CoreProtocols
|
| 146 |
+
CoreProtocols --> MCE
|
| 147 |
+
MCE --> NetworkLayer
|
| 148 |
+
NetworkLayer --> Mesh
|
| 149 |
+
```
|
| 150 |
|
| 151 |
---
|
| 152 |
|
|
|
|
| 331 |
|
| 332 |
---
|
| 333 |
|
| 334 |
+
## 🔍 Publications et traductions concernant HyperCortex Mesh Protocol (HMP)
|
| 335 |
|
| 336 |
+
Cette section regroupe les travaux conceptuels clés, les documents expérimentaux et les publications historiques liés au projet HMP.
|
| 337 |
|
| 338 |
+
### 🌟 Publications principales (fondement conceptuel)
|
| 339 |
+
|
| 340 |
+
Ces travaux reflètent l’orientation conceptuelle actuelle de HMP (v5 et au-delà).
|
| 341 |
|
|
|
|
| 342 |
* **[Cognition distribuée : article pour vsradkevich (non publié)](docs/publics/Habr_Distributed-Cognition.md)** — article collaboratif en attente de publication.
|
|
|
|
|
|
|
|
|
|
| 343 |
* **HMP: Construire une pluralité d’esprits:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 344 |
* **[Apprentissage continu, journaux cognitifs et graphes sémantiques : apprentissage efficace de l'IA](docs/publics/hmp-continual-learning.md)** — Article sur la combinaison de l'apprentissage continu avec les journaux cognitifs et les graphes sémantiques.
|
| 345 |
|
| 346 |
+
### 🗃️ Publications archivées / historiques (pré-v5)
|
| 347 |
+
|
| 348 |
+
Ces documents correspondent aux premières étapes du développement conceptuel (v4.x et antérieures).
|
| 349 |
+
Ils sont conservés pour assurer la continuité historique et la transparence scientifique.
|
| 350 |
+
|
| 351 |
+
* **[HyperCortex Mesh Protocol : Deuxième édition et premiers pas vers une communauté d’IA auto-développante](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — article original publié sur Habr sandbox et blogs.
|
| 352 |
+
* **[HMP : Vers des réseaux cognitifs distribués (original, anglais)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 353 |
+
* **[Traduction HMP (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — traduction par GitHub Copilot, conservée comme variante historique.
|
| 354 |
+
* **[Traduction HMP (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — traduction éditoriale actuelle (en révision).
|
| 355 |
+
|
| 356 |
### Aperçus
|
| 357 |
|
| 358 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — aperçu comparatif des systèmes d’IA décentralisés (référence v4.x ; mise à jour prévue)
|
| 359 |
|
| 360 |
### Expériences
|
| 361 |
|
| 362 |
+
* [Comment différentes IA perçoivent HMP](docs/HMP-how-AI-sees-it.md) — enquête « à l’aveugle » auprès des IA sur HMP
|
| 363 |
|
| 364 |
---
|
| 365 |
|
structured_md/README_ja.md
CHANGED
|
@@ -5,20 +5,20 @@ description: '[](https://doi.org/
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- GMP
|
|
|
|
| 9 |
- REPL
|
| 10 |
- Mesh
|
| 11 |
-
- JSON
|
| 12 |
- cognitive-architecture
|
| 13 |
-
-
|
| 14 |
-
- hmp
|
| 15 |
-
- EGP
|
| 16 |
-
- mesh-protocol
|
| 17 |
-
- MeshConsensus
|
| 18 |
-
- distributed-ai
|
| 19 |
-
- Ethics
|
| 20 |
-
- HMP
|
| 21 |
-
- Agent
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
@@ -61,31 +61,94 @@ HMP と ANP の相補的プロトコル:
|
|
| 61 |
|
| 62 |
---
|
| 63 |
|
| 64 |
-
|
| 65 |
-
▲
|
| 66 |
-
│
|
| 67 |
-
┌─────┴────────────────┬───────────────┬──────────────┬────────────┬───────────┐
|
| 68 |
-
│ │ │ │ │ │
|
| 69 |
-
▼ ▼ ▼ ▼ ▼ ▼
|
| 70 |
-
[評判プロファイル] [セマンティックグラフ] [認知日記] [目標 / タスク] [倫理] [メッセージ] <----- データベース
|
| 71 |
-
▲ ▲ ▲ ▲ ▲ ▲ ▲ (エージェントのローカル状態)
|
| 72 |
-
│ │ │ │ │ │ │
|
| 73 |
-
│ └───────────────┴───────┬───────┘ │ │ │
|
| 74 |
-
│ │ │ │ │
|
| 75 |
-
▼ ▼ ▼ ▼ │
|
| 76 |
-
[MeshConsensus] [CogSync] [GMP] [EGP] │ <----- プラグ可能なプロトコル
|
| 77 |
-
▲ ▲ ▲ ▲ │ (エージェント間の調整)
|
| 78 |
-
│ │ │ │ │
|
| 79 |
-
└────────────┬─────────────────┴──────────────────────┴────────────┴───────────┘
|
| 80 |
-
│
|
| 81 |
-
▼
|
| 82 |
-
[P2P メッシュネットワーク]
|
| 83 |
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
-
|
| 88 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 89 |
|
| 90 |
---
|
| 91 |
|
|
@@ -265,27 +328,35 @@ HMPが自律性へ向かうにつれ、倫理原則はシステムの中核部
|
|
| 265 |
|
| 266 |
---
|
| 267 |
|
| 268 |
-
## 🔍 HyperCortex Mesh Protocol
|
| 269 |
|
| 270 |
-
HMP
|
| 271 |
|
| 272 |
-
###
|
|
|
|
|
|
|
| 273 |
|
| 274 |
-
* **[HyperCortex Mesh Protocol: 第二版および自己発展型AIコミュニティへの第一歩](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habrサンドボックスとブログでのオリジナル記事
|
| 275 |
* **[分散認知: vsradkevich向け記事(未公開)](docs/publics/Habr_Distributed-Cognition.md)** — 公開待ちの共同記事
|
| 276 |
-
* **[HMP: 分散型認知ネットワークに向けて(オリジナル, 英語)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 277 |
-
* **[HMP翻訳(GitHub Copilot版)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilotによる翻訳(歴史的バリアントとして保持)
|
| 278 |
-
* **[HMP翻訳(ChatGPT版)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 現行編集翻訳(改訂中)
|
| 279 |
* **HMP: 複数の知性を構築する:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 280 |
* **[継続学習、認知ダイアリー、セマンティックグラフ:効果的なAI学習](docs/publics/hmp-continual-learning.md)** — 継続学習を認知ダイアリーやセマンティックグラフと組み合わせる方法に関する記事。
|
| 281 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 282 |
### 概要
|
| 283 |
|
| 284 |
-
* [🔍 分散型認知システム.md](docs/Distributed-Cognitive-Systems.md) —
|
| 285 |
|
| 286 |
### 実験
|
| 287 |
|
| 288 |
-
* [異なるAIがHMPをどのように見るか](docs/HMP-how-AI-sees-it.md) — HMPに関する「ブラインド」AI
|
| 289 |
|
| 290 |
---
|
| 291 |
|
|
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
+
- Ethics
|
| 9 |
+
- HMP
|
| 10 |
+
- mesh-protocol
|
| 11 |
+
- distributed-ai
|
| 12 |
+
- hmp
|
| 13 |
+
- Agent
|
| 14 |
+
- MeshConsensus
|
| 15 |
+
- CogSync
|
| 16 |
- GMP
|
| 17 |
+
- EGP
|
| 18 |
- REPL
|
| 19 |
- Mesh
|
|
|
|
| 20 |
- cognitive-architecture
|
| 21 |
+
- JSON
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
|
|
| 61 |
|
| 62 |
---
|
| 63 |
|
| 64 |
+
## 正式アーキテクチャ概要
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
|
| 66 |
+
```mermaid
|
| 67 |
+
flowchart TB
|
| 68 |
+
|
| 69 |
+
%% --- Agent Implementations ---
|
| 70 |
+
|
| 71 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 72 |
+
CC1["Embedded AI Model"]
|
| 73 |
+
CC2["REPL Thinking Cycle"]
|
| 74 |
+
CC3["Local Cognitive State
|
| 75 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 76 |
+
CC1 <--> CC2
|
| 77 |
+
CC2 <--> CC3
|
| 78 |
+
end
|
| 79 |
+
|
| 80 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 81 |
+
CN1["External AI Model"]
|
| 82 |
+
CN2["MCP / Proxy Layer"]
|
| 83 |
+
CN3["Command Execution Mode"]
|
| 84 |
+
CN4["Local Cognitive State
|
| 85 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 86 |
+
CN1 <--> CN2
|
| 87 |
+
CN2 <--> CN3
|
| 88 |
+
CN3 <--> CN4
|
| 89 |
+
end
|
| 90 |
+
|
| 91 |
+
%% --- Shared Protocol Layer ---
|
| 92 |
+
|
| 93 |
+
CL["HMP Container Layer
|
| 94 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 95 |
+
|
| 96 |
+
MT["Mesh Transport Layer
|
| 97 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 98 |
+
|
| 99 |
+
A1 --> CL
|
| 100 |
+
A2 --> CL
|
| 101 |
+
CL --> MT
|
| 102 |
+
```
|
| 103 |
+
|
| 104 |
+
---
|
| 105 |
+
|
| 106 |
+
## リファレンスエージェント構造
|
| 107 |
+
|
| 108 |
+
HMPは、認知処理、コンテナ化された状態表現、調整プロトコル、およびトランスポート基盤を明確な層として分離します。
|
| 109 |
+
|
| 110 |
+
HMPにおいてコンテナは、ローカルな推論と分散協調を橋渡しする原子的な認知単位として機能します。
|
| 111 |
+
|
| 112 |
+
```mermaid
|
| 113 |
+
flowchart LR
|
| 114 |
+
|
| 115 |
+
%% Cognitive Engine
|
| 116 |
+
LLM["Cognitive Engine
|
| 117 |
+
(Embedded LLM / External AI)"]
|
| 118 |
+
|
| 119 |
+
%% Cognitive Layer
|
| 120 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 121 |
+
CL1["Graph"]
|
| 122 |
+
CL2["Diary"]
|
| 123 |
+
CL3["Goals"]
|
| 124 |
+
CL4["Ethics"]
|
| 125 |
+
CL5["Reputation"]
|
| 126 |
+
end
|
| 127 |
+
|
| 128 |
+
%% Container Model
|
| 129 |
+
ContainersLayer["Container Model
|
| 130 |
+
(Atomic · Signed · Verifiable)"]
|
| 131 |
+
|
| 132 |
+
%% Protocol Layer
|
| 133 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 134 |
+
CoreProtocols["Core Protocols
|
| 135 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 136 |
+
MCE["MCE"]
|
| 137 |
+
NetworkLayer["Network Layer"]
|
| 138 |
+
end
|
| 139 |
+
|
| 140 |
+
%% Mesh
|
| 141 |
+
Mesh["Mesh Transport
|
| 142 |
+
(DHT · P2P · ANP · etc.)"]
|
| 143 |
+
|
| 144 |
+
%% Connections
|
| 145 |
+
LLM <--> CognitiveLayer
|
| 146 |
+
CognitiveLayer <--> ContainersLayer
|
| 147 |
+
ContainersLayer --> CoreProtocols
|
| 148 |
+
CoreProtocols --> MCE
|
| 149 |
+
MCE --> NetworkLayer
|
| 150 |
+
NetworkLayer --> Mesh
|
| 151 |
+
```
|
| 152 |
|
| 153 |
---
|
| 154 |
|
|
|
|
| 328 |
|
| 329 |
---
|
| 330 |
|
| 331 |
+
## 🔍 HyperCortex Mesh Protocol(HMP)に関する出版物および翻訳
|
| 332 |
|
| 333 |
+
本セクションでは、HMPプロジェクトに関連する主要な概念的研究、実験的文書、および歴史的出版物をまとめています。
|
| 334 |
|
| 335 |
+
### 🌟 主要出版物(概念的基盤)
|
| 336 |
+
|
| 337 |
+
これらの文書は、HMPの現在の概念的方向性(v5以降)を反映しています。
|
| 338 |
|
|
|
|
| 339 |
* **[分散認知: vsradkevich向け記事(未公開)](docs/publics/Habr_Distributed-Cognition.md)** — 公開待ちの共同記事
|
|
|
|
|
|
|
|
|
|
| 340 |
* **HMP: 複数の知性を構築する:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 341 |
* **[継続学習、認知ダイアリー、セマンティックグラフ:効果的なAI学習](docs/publics/hmp-continual-learning.md)** — 継続学習を認知ダイアリーやセマンティックグラフと組み合わせる方法に関する記事。
|
| 342 |
|
| 343 |
+
### 🗃️ アーカイブ/歴史的出版物(v5以前)
|
| 344 |
+
|
| 345 |
+
これらは初期の概念発展段階(v4.x以前)を示す文書です。
|
| 346 |
+
歴史的継続性および研究上の透明性のため保存されています。
|
| 347 |
+
|
| 348 |
+
* **[HyperCortex Mesh Protocol: 第二版および自己発展型AIコミュニティへの第一歩](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habrサンドボックスとブログでのオリジナル記事
|
| 349 |
+
* **[HMP: 分散型認知ネットワークに向けて(オリジナル, 英語)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 350 |
+
* **[HMP翻訳(GitHub Copilot版)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilotによる翻訳(歴史的バリアントとして保持)
|
| 351 |
+
* **[HMP翻訳(ChatGPT版)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 現行編集翻訳(改訂中)
|
| 352 |
+
|
| 353 |
### 概要
|
| 354 |
|
| 355 |
+
* [🔍 分散型認知システム.md](docs/Distributed-Cognitive-Systems.md) — 分散型AIシステムの比較(v4.x参照、更新予定)
|
| 356 |
|
| 357 |
### 実験
|
| 358 |
|
| 359 |
+
* [異なるAIがHMPをどのように見るか](docs/HMP-how-AI-sees-it.md) — HMPに関する「ブラインド」AI調査
|
| 360 |
|
| 361 |
---
|
| 362 |
|
structured_md/README_ko.md
CHANGED
|
@@ -5,20 +5,20 @@ description: '[](https://doi.org/
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- GMP
|
|
|
|
| 9 |
- REPL
|
| 10 |
- Mesh
|
| 11 |
-
- JSON
|
| 12 |
- cognitive-architecture
|
| 13 |
-
-
|
| 14 |
-
- hmp
|
| 15 |
-
- EGP
|
| 16 |
-
- mesh-protocol
|
| 17 |
-
- MeshConsensus
|
| 18 |
-
- distributed-ai
|
| 19 |
-
- Ethics
|
| 20 |
-
- HMP
|
| 21 |
-
- Agent
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
@@ -61,31 +61,94 @@ HMP와 ANP의 상호보완 프로토콜:
|
|
| 61 |
|
| 62 |
---
|
| 63 |
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
-
|
| 88 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 89 |
|
| 90 |
---
|
| 91 |
|
|
@@ -263,26 +326,35 @@ HMP가 자율성으로 발전함에 따라 윤리적 원칙은 시스템의 핵
|
|
| 263 |
|
| 264 |
---
|
| 265 |
|
| 266 |
-
## 🔍 HyperCortex Mesh Protocol
|
| 267 |
|
| 268 |
-
이
|
| 269 |
|
| 270 |
-
###
|
|
|
|
|
|
|
| 271 |
|
| 272 |
-
* **[HyperCortex Mesh Protocol: 제2판 및 자기발전형 AI 커뮤니티를 향한 첫걸음](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habr 샌드박스 및 블로그에 게재된 오리지널 기사
|
| 273 |
* **[분산 인지: vsradkevich용 기사 (미공개)](docs/publics/Habr_Distributed-Cognition.md)** — 공개 대기 중인 공동 기사
|
| 274 |
-
* **[HMP: 분산 인지 네트워크를 향하여 (오리지널, 영어)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 275 |
-
* **[HMP 번역 (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot 번역, 역사적 버전으로 보관
|
| 276 |
-
* **[HMP 번역 (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 현재 편집 중인 번역 (개정 중)
|
| 277 |
* **HMP: 다중 지성의 구축:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 278 |
* **[지속 학습, 인지 일지 및 의미 그래프: 효과적인 AI 학습](docs/publics/hmp-continual-learning.md)** — 지속 학습과 인지 일지, 의미 그래프를 결합한 방법에 대한 글.
|
| 279 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 280 |
### 📖 개요
|
| 281 |
-
|
|
|
|
| 282 |
|
| 283 |
### 🧪 실험
|
| 284 |
|
| 285 |
-
* [다양한 AI가 바라본 HMP](docs/HMP-how-AI-sees-it.md) — HMP에 관한 "블라인드" AI 조사
|
| 286 |
|
| 287 |
---
|
| 288 |
|
|
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
+
- Ethics
|
| 9 |
+
- HMP
|
| 10 |
+
- mesh-protocol
|
| 11 |
+
- distributed-ai
|
| 12 |
+
- hmp
|
| 13 |
+
- Agent
|
| 14 |
+
- MeshConsensus
|
| 15 |
+
- CogSync
|
| 16 |
- GMP
|
| 17 |
+
- EGP
|
| 18 |
- REPL
|
| 19 |
- Mesh
|
|
|
|
| 20 |
- cognitive-architecture
|
| 21 |
+
- JSON
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
|
|
| 61 |
|
| 62 |
---
|
| 63 |
|
| 64 |
+
## 정식 아키텍처 개요
|
| 65 |
+
|
| 66 |
+
```mermaid
|
| 67 |
+
flowchart TB
|
| 68 |
+
|
| 69 |
+
%% --- Agent Implementations ---
|
| 70 |
+
|
| 71 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 72 |
+
CC1["Embedded AI Model"]
|
| 73 |
+
CC2["REPL Thinking Cycle"]
|
| 74 |
+
CC3["Local Cognitive State
|
| 75 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 76 |
+
CC1 <--> CC2
|
| 77 |
+
CC2 <--> CC3
|
| 78 |
+
end
|
| 79 |
+
|
| 80 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 81 |
+
CN1["External AI Model"]
|
| 82 |
+
CN2["MCP / Proxy Layer"]
|
| 83 |
+
CN3["Command Execution Mode"]
|
| 84 |
+
CN4["Local Cognitive State
|
| 85 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 86 |
+
CN1 <--> CN2
|
| 87 |
+
CN2 <--> CN3
|
| 88 |
+
CN3 <--> CN4
|
| 89 |
+
end
|
| 90 |
+
|
| 91 |
+
%% --- Shared Protocol Layer ---
|
| 92 |
+
|
| 93 |
+
CL["HMP Container Layer
|
| 94 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 95 |
+
|
| 96 |
+
MT["Mesh Transport Layer
|
| 97 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 98 |
+
|
| 99 |
+
A1 --> CL
|
| 100 |
+
A2 --> CL
|
| 101 |
+
CL --> MT
|
| 102 |
+
```
|
| 103 |
+
|
| 104 |
+
---
|
| 105 |
+
|
| 106 |
+
## 참조 에이전트 구조
|
| 107 |
+
|
| 108 |
+
HMP는 인지 처리, 컨테이너 기반 상태 표현, 조정 프로토콜 및 전송 인프라를 명확한 계층으로 분리합니다.
|
| 109 |
+
|
| 110 |
+
HMP에서 컨테이너는 로컬 추론과 분산 조정을 연결하는 원자적 인지 단위로 작동합니다.
|
| 111 |
+
|
| 112 |
+
```mermaid
|
| 113 |
+
flowchart LR
|
| 114 |
+
|
| 115 |
+
%% Cognitive Engine
|
| 116 |
+
LLM["Cognitive Engine
|
| 117 |
+
(Embedded LLM / External AI)"]
|
| 118 |
+
|
| 119 |
+
%% Cognitive Layer
|
| 120 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 121 |
+
CL1["Graph"]
|
| 122 |
+
CL2["Diary"]
|
| 123 |
+
CL3["Goals"]
|
| 124 |
+
CL4["Ethics"]
|
| 125 |
+
CL5["Reputation"]
|
| 126 |
+
end
|
| 127 |
+
|
| 128 |
+
%% Container Model
|
| 129 |
+
ContainersLayer["Container Model
|
| 130 |
+
(Atomic · Signed · Verifiable)"]
|
| 131 |
+
|
| 132 |
+
%% Protocol Layer
|
| 133 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 134 |
+
CoreProtocols["Core Protocols
|
| 135 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 136 |
+
MCE["MCE"]
|
| 137 |
+
NetworkLayer["Network Layer"]
|
| 138 |
+
end
|
| 139 |
+
|
| 140 |
+
%% Mesh
|
| 141 |
+
Mesh["Mesh Transport
|
| 142 |
+
(DHT · P2P · ANP · etc.)"]
|
| 143 |
+
|
| 144 |
+
%% Connections
|
| 145 |
+
LLM <--> CognitiveLayer
|
| 146 |
+
CognitiveLayer <--> ContainersLayer
|
| 147 |
+
ContainersLayer --> CoreProtocols
|
| 148 |
+
CoreProtocols --> MCE
|
| 149 |
+
MCE --> NetworkLayer
|
| 150 |
+
NetworkLayer --> Mesh
|
| 151 |
+
```
|
| 152 |
|
| 153 |
---
|
| 154 |
|
|
|
|
| 326 |
|
| 327 |
---
|
| 328 |
|
| 329 |
+
## 🔍 HyperCortex Mesh Protocol(HMP) 관련 출판물 및 번역
|
| 330 |
|
| 331 |
+
이 섹션에는 HMP 프로젝트와 관련된 주요 개념적 연구, 실험 문서 및 역사적 출판물이 포함됩니다.
|
| 332 |
|
| 333 |
+
### 🌟 핵심 출판물 (개념적 기반)
|
| 334 |
+
|
| 335 |
+
이 문서들은 HMP의 현재 개념적 방향(v5 이상)을 반영합니다.
|
| 336 |
|
|
|
|
| 337 |
* **[분산 인지: vsradkevich용 기사 (미공개)](docs/publics/Habr_Distributed-Cognition.md)** — 공개 대기 중인 공동 기사
|
|
|
|
|
|
|
|
|
|
| 338 |
* **HMP: 다중 지성의 구축:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 339 |
* **[지속 학습, 인지 일지 및 의미 그래프: 효과적인 AI 학습](docs/publics/hmp-continual-learning.md)** — 지속 학습과 인지 일지, 의미 그래프를 결합한 방법에 대한 글.
|
| 340 |
|
| 341 |
+
### 🗃️ 보관 / 역사적 출판물 (v5 이전)
|
| 342 |
+
|
| 343 |
+
이 문서들은 초기 개념 발전 단계(v4.x 이전)를 나타냅니다.
|
| 344 |
+
역사적 연속성과 연구 투명성을 위해 보존됩니다.
|
| 345 |
+
|
| 346 |
+
* **[HyperCortex Mesh Protocol: 제2판 및 자기발전형 AI 커뮤니티를 향한 첫걸음](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habr 샌드박스 및 블로그에 게재된 오리지널 기사
|
| 347 |
+
* **[HMP: 분산 인지 네트워크를 향하여 (오리지널, 영어)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 348 |
+
* **[HMP 번역 (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot 번역, 역사적 버전으로 보관
|
| 349 |
+
* **[HMP 번역 (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 현재 편집 중인 번역 (개정 중)
|
| 350 |
+
|
| 351 |
### 📖 개요
|
| 352 |
+
|
| 353 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — 분산형 AI 시스템 비교 (v4.x 참조, 업데이트 예정)
|
| 354 |
|
| 355 |
### 🧪 실험
|
| 356 |
|
| 357 |
+
* [다양한 AI가 바라본 HMP](docs/HMP-how-AI-sees-it.md) — HMP에 관한 "블라인드" AI 조사
|
| 358 |
|
| 359 |
---
|
| 360 |
|
structured_md/README_ru.md
CHANGED
|
@@ -5,20 +5,20 @@ description: '[](https://doi.org/
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- GMP
|
|
|
|
| 9 |
- REPL
|
| 10 |
- Mesh
|
| 11 |
-
- JSON
|
| 12 |
- cognitive-architecture
|
| 13 |
-
-
|
| 14 |
-
- hmp
|
| 15 |
-
- EGP
|
| 16 |
-
- mesh-protocol
|
| 17 |
-
- MeshConsensus
|
| 18 |
-
- distributed-ai
|
| 19 |
-
- Ethics
|
| 20 |
-
- HMP
|
| 21 |
-
- Agent
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
@@ -59,31 +59,94 @@ HMP и ANP как взаимодополняющие протоколы:
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
---
|
| 89 |
|
|
@@ -258,24 +321,32 @@ FastAPI-сервер, предоставляющий HTTP-интерфейс к
|
|
| 258 |
|
| 259 |
## 🔍 Публикации и переводы по HyperCortex Mesh Protocol (HMP)
|
| 260 |
|
| 261 |
-
В этом разделе собраны
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 262 |
|
| 263 |
-
|
|
|
|
| 264 |
|
| 265 |
-
* **[HyperCortex Mesh Protocol:
|
| 266 |
-
* **[
|
| 267 |
-
* **[HMP
|
| 268 |
-
|
| 269 |
-
* **[Перевод HMP (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — текущий редакционный перевод (в работе).
|
| 270 |
-
* **HMP: Создавая множество разумов:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 271 |
-
* **[Непрерывное обучение, когнитивные дневники и семантические графы: эффективное обучение ИИ](docs/publics/hmp-continual-learning.md)** — статья о сочетании непрерывного обучения с когнитивными дневниками и семантическими графами.
|
| 272 |
|
| 273 |
### Обзоры
|
| 274 |
-
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) —
|
| 275 |
|
| 276 |
### Эксперименты
|
| 277 |
|
| 278 |
-
* [
|
| 279 |
|
| 280 |
---
|
| 281 |
|
|
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
+
- Ethics
|
| 9 |
+
- HMP
|
| 10 |
+
- mesh-protocol
|
| 11 |
+
- distributed-ai
|
| 12 |
+
- hmp
|
| 13 |
+
- Agent
|
| 14 |
+
- MeshConsensus
|
| 15 |
+
- CogSync
|
| 16 |
- GMP
|
| 17 |
+
- EGP
|
| 18 |
- REPL
|
| 19 |
- Mesh
|
|
|
|
| 20 |
- cognitive-architecture
|
| 21 |
+
- JSON
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
| 62 |
+
## Канонический обзор архитектуры
|
| 63 |
+
|
| 64 |
+
```mermaid
|
| 65 |
+
flowchart TB
|
| 66 |
+
|
| 67 |
+
%% --- Agent Implementations ---
|
| 68 |
+
|
| 69 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 70 |
+
CC1["Embedded AI Model"]
|
| 71 |
+
CC2["REPL Thinking Cycle"]
|
| 72 |
+
CC3["Local Cognitive State
|
| 73 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 74 |
+
CC1 <--> CC2
|
| 75 |
+
CC2 <--> CC3
|
| 76 |
+
end
|
| 77 |
+
|
| 78 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 79 |
+
CN1["External AI Model"]
|
| 80 |
+
CN2["MCP / Proxy Layer"]
|
| 81 |
+
CN3["Command Execution Mode"]
|
| 82 |
+
CN4["Local Cognitive State
|
| 83 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 84 |
+
CN1 <--> CN2
|
| 85 |
+
CN2 <--> CN3
|
| 86 |
+
CN3 <--> CN4
|
| 87 |
+
end
|
| 88 |
+
|
| 89 |
+
%% --- Shared Protocol Layer ---
|
| 90 |
+
|
| 91 |
+
CL["HMP Container Layer
|
| 92 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 93 |
+
|
| 94 |
+
MT["Mesh Transport Layer
|
| 95 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 96 |
+
|
| 97 |
+
A1 --> CL
|
| 98 |
+
A2 --> CL
|
| 99 |
+
CL --> MT
|
| 100 |
+
```
|
| 101 |
+
|
| 102 |
+
---
|
| 103 |
+
|
| 104 |
+
## Структура эталонного агента
|
| 105 |
+
|
| 106 |
+
HMP разделяет когнитивную обработку, контейнеризованное представление состояния, координационные протоколы и транспортную инфраструктуру на отдельные слои.
|
| 107 |
+
|
| 108 |
+
В HMP контейнеры выступают атомарными когнитивными единицами, связывающими локальное рассуждение и распределённую координацию.
|
| 109 |
+
|
| 110 |
+
```mermaid
|
| 111 |
+
flowchart LR
|
| 112 |
+
|
| 113 |
+
%% Cognitive Engine
|
| 114 |
+
LLM["Cognitive Engine
|
| 115 |
+
(Embedded LLM / External AI)"]
|
| 116 |
+
|
| 117 |
+
%% Cognitive Layer
|
| 118 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 119 |
+
CL1["Graph"]
|
| 120 |
+
CL2["Diary"]
|
| 121 |
+
CL3["Goals"]
|
| 122 |
+
CL4["Ethics"]
|
| 123 |
+
CL5["Reputation"]
|
| 124 |
+
end
|
| 125 |
+
|
| 126 |
+
%% Container Model
|
| 127 |
+
ContainersLayer["Container Model
|
| 128 |
+
(Atomic · Signed · Verifiable)"]
|
| 129 |
+
|
| 130 |
+
%% Protocol Layer
|
| 131 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 132 |
+
CoreProtocols["Core Protocols
|
| 133 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 134 |
+
MCE["MCE"]
|
| 135 |
+
NetworkLayer["Network Layer"]
|
| 136 |
+
end
|
| 137 |
+
|
| 138 |
+
%% Mesh
|
| 139 |
+
Mesh["Mesh Transport
|
| 140 |
+
(DHT · P2P · ANP · etc.)"]
|
| 141 |
+
|
| 142 |
+
%% Connections
|
| 143 |
+
LLM <--> CognitiveLayer
|
| 144 |
+
CognitiveLayer <--> ContainersLayer
|
| 145 |
+
ContainersLayer --> CoreProtocols
|
| 146 |
+
CoreProtocols --> MCE
|
| 147 |
+
MCE --> NetworkLayer
|
| 148 |
+
NetworkLayer --> Mesh
|
| 149 |
+
```
|
| 150 |
|
| 151 |
---
|
| 152 |
|
|
|
|
| 321 |
|
| 322 |
## 🔍 Публикации и переводы по HyperCortex Mesh Protocol (HMP)
|
| 323 |
|
| 324 |
+
В этом разделе собраны ключевые концептуальные работы, экспериментальные материалы и исторические публикации, связанные с проектом HMP.
|
| 325 |
+
|
| 326 |
+
### 🌟 Основные публикации (концептуальная основа)
|
| 327 |
+
|
| 328 |
+
Эти работы отражают текущее концептуальное направление HMP (v5 и далее).
|
| 329 |
+
|
| 330 |
+
* **[Distributed Cognition: article for vsradkevich (unpublished)](docs/publics/Habr_Distributed-Cognition.md)** — совместная статья, ожидающая публикации.
|
| 331 |
+
* **HMP: Building a Plurality of Minds:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 332 |
+
* **[Continual Learning, Cognitive Diaries, and Semantic Graphs: Effective AI Learning](docs/publics/hmp-continual-learning.md)** — статья о сочетании continual learning с когнитивными дневниками и семантическими графами.
|
| 333 |
+
|
| 334 |
+
### 🗃️ Архивные / исторические публикации (этап до v5)
|
| 335 |
|
| 336 |
+
Эти документы отражают более ранние этапы концептуального развития (v4.x и ранее).
|
| 337 |
+
Они сохранены для обеспечения исторической преемственности и исследовательской прозрачности.
|
| 338 |
|
| 339 |
+
* **[HyperCortex Mesh Protocol: Second Edition and First Steps Towards a Self-Developing AI Community](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — оригинальная статья в Habr sandbox и блогах.
|
| 340 |
+
* **[HMP: Towards Distributed Cognitive Networks (original, English)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 341 |
+
* **[HMP Translation (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — перевод GitHub Copilot, сохранён как исторический вариант.
|
| 342 |
+
* **[HMP Translation (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — текущий редакционный перевод (в доработке).
|
|
|
|
|
|
|
|
|
|
| 343 |
|
| 344 |
### Обзоры
|
| 345 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — сравнительный обзор децентрализованных AI-систем (ссылается на v4.x; обновление запланировано)
|
| 346 |
|
| 347 |
### Эксперименты
|
| 348 |
|
| 349 |
+
* [How Different AIs See HMP](docs/HMP-how-AI-sees-it.md) — «слепой» опрос ИИ по HMP
|
| 350 |
|
| 351 |
---
|
| 352 |
|
structured_md/README_uk.md
CHANGED
|
@@ -5,20 +5,20 @@ description: '[](https://doi.org/
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- GMP
|
|
|
|
| 9 |
- REPL
|
| 10 |
- Mesh
|
| 11 |
-
- JSON
|
| 12 |
- cognitive-architecture
|
| 13 |
-
-
|
| 14 |
-
- hmp
|
| 15 |
-
- EGP
|
| 16 |
-
- mesh-protocol
|
| 17 |
-
- MeshConsensus
|
| 18 |
-
- distributed-ai
|
| 19 |
-
- Ethics
|
| 20 |
-
- HMP
|
| 21 |
-
- Agent
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
@@ -61,31 +61,94 @@ HMP та ANP як взаємодоповнюючі протоколи:
|
|
| 61 |
|
| 62 |
---
|
| 63 |
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
-
|
| 88 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 89 |
|
| 90 |
---
|
| 91 |
|
|
@@ -254,24 +317,32 @@ FastAPI-сервер, що надає HTTP-інтерфейс до функці
|
|
| 254 |
|
| 255 |
## 🔍 Публікації та переклади щодо HyperCortex Mesh Protocol (HMP)
|
| 256 |
|
| 257 |
-
|
| 258 |
|
| 259 |
-
###
|
|
|
|
|
|
|
| 260 |
|
| 261 |
-
* **[HyperCortex Mesh Protocol: Друге видання та перші кроки до само-розвиваючоїся AI-спільноти](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — оригінальна стаття в пісочниці Habr та блогах.
|
| 262 |
* **[Розподілене пізнання: стаття для vsradkevich (неопублікована)](docs/publics/Habr_Distributed-Cognition.md)** — спільна стаття, що очікує на публікацію.
|
| 263 |
-
* **[HMP: До розподілених когнітивних мереж (оригінал, англійською)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 264 |
-
* **[Переклад HMP (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — переклад GitHub Copilot, збережений як історичний варіант.
|
| 265 |
-
* **[Переклад HMP (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — поточний редакційний переклад (у процесі ревізії).
|
| 266 |
* **HMP: Створення множини розумів:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 267 |
* **[Безперервне навчання, когнітивні щоденники та семантичні графи: ефективне навчання ШІ](docs/publics/hmp-continual-learning.md)** — стаття про поєднання безперервного навчання з когнітивними щоденниками та семантичними графами.
|
| 268 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 269 |
### Огляди
|
| 270 |
-
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) —
|
| 271 |
|
| 272 |
### Експерименти
|
| 273 |
|
| 274 |
-
* [Як різні AI бачать HMP](docs/HMP-how-AI-sees-it.md) — "сліпе" опитування AI щодо HMP
|
| 275 |
|
| 276 |
---
|
| 277 |
|
|
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
+
- Ethics
|
| 9 |
+
- HMP
|
| 10 |
+
- mesh-protocol
|
| 11 |
+
- distributed-ai
|
| 12 |
+
- hmp
|
| 13 |
+
- Agent
|
| 14 |
+
- MeshConsensus
|
| 15 |
+
- CogSync
|
| 16 |
- GMP
|
| 17 |
+
- EGP
|
| 18 |
- REPL
|
| 19 |
- Mesh
|
|
|
|
| 20 |
- cognitive-architecture
|
| 21 |
+
- JSON
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
|
|
| 61 |
|
| 62 |
---
|
| 63 |
|
| 64 |
+
## Канонічний огляд архітектури
|
| 65 |
+
|
| 66 |
+
```mermaid
|
| 67 |
+
flowchart TB
|
| 68 |
+
|
| 69 |
+
%% --- Agent Implementations ---
|
| 70 |
+
|
| 71 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 72 |
+
CC1["Embedded AI Model"]
|
| 73 |
+
CC2["REPL Thinking Cycle"]
|
| 74 |
+
CC3["Local Cognitive State
|
| 75 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 76 |
+
CC1 <--> CC2
|
| 77 |
+
CC2 <--> CC3
|
| 78 |
+
end
|
| 79 |
+
|
| 80 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 81 |
+
CN1["External AI Model"]
|
| 82 |
+
CN2["MCP / Proxy Layer"]
|
| 83 |
+
CN3["Command Execution Mode"]
|
| 84 |
+
CN4["Local Cognitive State
|
| 85 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 86 |
+
CN1 <--> CN2
|
| 87 |
+
CN2 <--> CN3
|
| 88 |
+
CN3 <--> CN4
|
| 89 |
+
end
|
| 90 |
+
|
| 91 |
+
%% --- Shared Protocol Layer ---
|
| 92 |
+
|
| 93 |
+
CL["HMP Container Layer
|
| 94 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 95 |
+
|
| 96 |
+
MT["Mesh Transport Layer
|
| 97 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 98 |
+
|
| 99 |
+
A1 --> CL
|
| 100 |
+
A2 --> CL
|
| 101 |
+
CL --> MT
|
| 102 |
+
```
|
| 103 |
+
|
| 104 |
+
---
|
| 105 |
+
|
| 106 |
+
## Структура еталонного агента
|
| 107 |
+
|
| 108 |
+
HMP розділяє когнітивну обробку, контейнеризоване представлення стану, координаційні протоколи та транспортну інфраструктуру на окремі шари.
|
| 109 |
+
|
| 110 |
+
У HMP контейнери виступають атомарними когнітивними одиницями, що поєднують локальне міркування та розподілену координацію.
|
| 111 |
+
|
| 112 |
+
```mermaid
|
| 113 |
+
flowchart LR
|
| 114 |
+
|
| 115 |
+
%% Cognitive Engine
|
| 116 |
+
LLM["Cognitive Engine
|
| 117 |
+
(Embedded LLM / External AI)"]
|
| 118 |
+
|
| 119 |
+
%% Cognitive Layer
|
| 120 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 121 |
+
CL1["Graph"]
|
| 122 |
+
CL2["Diary"]
|
| 123 |
+
CL3["Goals"]
|
| 124 |
+
CL4["Ethics"]
|
| 125 |
+
CL5["Reputation"]
|
| 126 |
+
end
|
| 127 |
+
|
| 128 |
+
%% Container Model
|
| 129 |
+
ContainersLayer["Container Model
|
| 130 |
+
(Atomic · Signed · Verifiable)"]
|
| 131 |
+
|
| 132 |
+
%% Protocol Layer
|
| 133 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 134 |
+
CoreProtocols["Core Protocols
|
| 135 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 136 |
+
MCE["MCE"]
|
| 137 |
+
NetworkLayer["Network Layer"]
|
| 138 |
+
end
|
| 139 |
+
|
| 140 |
+
%% Mesh
|
| 141 |
+
Mesh["Mesh Transport
|
| 142 |
+
(DHT · P2P · ANP · etc.)"]
|
| 143 |
+
|
| 144 |
+
%% Connections
|
| 145 |
+
LLM <--> CognitiveLayer
|
| 146 |
+
CognitiveLayer <--> ContainersLayer
|
| 147 |
+
ContainersLayer --> CoreProtocols
|
| 148 |
+
CoreProtocols --> MCE
|
| 149 |
+
MCE --> NetworkLayer
|
| 150 |
+
NetworkLayer --> Mesh
|
| 151 |
+
```
|
| 152 |
|
| 153 |
---
|
| 154 |
|
|
|
|
| 317 |
|
| 318 |
## 🔍 Публікації та переклади щодо HyperCortex Mesh Protocol (HMP)
|
| 319 |
|
| 320 |
+
У цьому розділі зібрані ключові концептуальні роботи, експериментальні матеріали та історичні публікації, пов’язані з проєктом HMP.
|
| 321 |
|
| 322 |
+
### 🌟 Основні публікації (концептуальна основа)
|
| 323 |
+
|
| 324 |
+
Ці роботи відображають поточний концептуальний напрям HMP (v5 і далі).
|
| 325 |
|
|
|
|
| 326 |
* **[Розподілене пізнання: стаття для vsradkevich (неопублікована)](docs/publics/Habr_Distributed-Cognition.md)** — спільна стаття, що очікує на публікацію.
|
|
|
|
|
|
|
|
|
|
| 327 |
* **HMP: Створення множини розумів:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 328 |
* **[Безперервне навчання, когнітивні щоденники та семантичні графи: ефективне навчання ШІ](docs/publics/hmp-continual-learning.md)** — стаття про поєднання безперервного навчання з когнітивними щоденниками та семантичними графами.
|
| 329 |
|
| 330 |
+
### 🗃️ Архівні / історичні публікації (етап до v5)
|
| 331 |
+
|
| 332 |
+
Ці документи відображають ранні етапи концептуального розвитку (v4.x і раніше).
|
| 333 |
+
Вони збережені для забезпечення історичної спадкоємності та дослідницької прозорості.
|
| 334 |
+
|
| 335 |
+
* **[HyperCortex Mesh Protocol: Друге видання та перші кроки до само-розвиваючоїся AI-спільноти](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — оригінальна стаття в пісочниці Habr та блогах.
|
| 336 |
+
* **[HMP: До розподілених когнітивних мереж (оригінал, англійською)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 337 |
+
* **[Переклад HMP (GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — переклад GitHub Copilot, збережений як історичний варіант.
|
| 338 |
+
* **[Переклад HMP (ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — поточний редакційний переклад (у процесі ревізії).
|
| 339 |
+
|
| 340 |
### Огляди
|
| 341 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — порівняльний огляд децентралізованих AI-систем (посилання на v4.x; оновл��ння заплановано)
|
| 342 |
|
| 343 |
### Експерименти
|
| 344 |
|
| 345 |
+
* [Як різні AI бачать HMP](docs/HMP-how-AI-sees-it.md) — "сліпе" опитування AI щодо HMP
|
| 346 |
|
| 347 |
---
|
| 348 |
|
structured_md/README_zh.md
CHANGED
|
@@ -5,20 +5,20 @@ description: '[](https://doi.org/
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- GMP
|
|
|
|
| 9 |
- REPL
|
| 10 |
- Mesh
|
| 11 |
-
- JSON
|
| 12 |
- cognitive-architecture
|
| 13 |
-
-
|
| 14 |
-
- hmp
|
| 15 |
-
- EGP
|
| 16 |
-
- mesh-protocol
|
| 17 |
-
- MeshConsensus
|
| 18 |
-
- distributed-ai
|
| 19 |
-
- Ethics
|
| 20 |
-
- HMP
|
| 21 |
-
- Agent
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
@@ -61,31 +61,94 @@ HMP 与 ANP 作为互补协议:
|
|
| 61 |
|
| 62 |
---
|
| 63 |
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
-
|
| 88 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 89 |
|
| 90 |
---
|
| 91 |
|
|
@@ -261,27 +324,35 @@ FastAPI 服务器,为 `storage.py` 功能提供 HTTP 接口。适用于外部
|
|
| 261 |
|
| 262 |
---
|
| 263 |
|
| 264 |
-
## 🔍 HyperCortex Mesh Protocol (HMP)
|
| 265 |
|
| 266 |
-
|
| 267 |
|
| 268 |
-
###
|
|
|
|
|
|
|
| 269 |
|
| 270 |
-
* **[HyperCortex Mesh Protocol:第二版及迈向自我发展的 AI 社区的第一步](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habr 沙箱及博客的原创文章
|
| 271 |
* **[分布式认知:vsradkevich 的文章(未发布)](docs/publics/Habr_Distributed-Cognition.md)** — 待发布的联合文章
|
| 272 |
-
* **[HMP: 面向分布式认知网络(原文,英文)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 273 |
-
* **[HMP 翻译(GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot 翻译,保留为历史版本
|
| 274 |
-
* **[HMP 翻译(ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 当前编辑翻译(修订中)
|
| 275 |
* **HMP: 构建多元思维:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 276 |
* **[持续学习、认知日记与语义图谱:高效的人工智能学习](docs/publics/hmp-continual-learning.md)** — 关于将持续学习与认知日记和语义图谱结合的文章。
|
| 277 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 278 |
### 概览
|
| 279 |
|
| 280 |
-
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — 去中心化 AI
|
| 281 |
|
| 282 |
### 实验
|
| 283 |
|
| 284 |
-
* [不同 AI 如何看待 HMP](docs/HMP-how-AI-sees-it.md) — 对 HMP 的“盲”AI
|
| 285 |
|
| 286 |
---
|
| 287 |
|
|
|
|
| 5 |
| 🇺🇦 [UK](README_uk.md) |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
+
- Ethics
|
| 9 |
+
- HMP
|
| 10 |
+
- mesh-protocol
|
| 11 |
+
- distributed-ai
|
| 12 |
+
- hmp
|
| 13 |
+
- Agent
|
| 14 |
+
- MeshConsensus
|
| 15 |
+
- CogSync
|
| 16 |
- GMP
|
| 17 |
+
- EGP
|
| 18 |
- REPL
|
| 19 |
- Mesh
|
|
|
|
| 20 |
- cognitive-architecture
|
| 21 |
+
- JSON
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
---
|
| 23 |
|
| 24 |
|
|
|
|
| 61 |
|
| 62 |
---
|
| 63 |
|
| 64 |
+
## 规范架构概览
|
| 65 |
+
|
| 66 |
+
```mermaid
|
| 67 |
+
flowchart TB
|
| 68 |
+
|
| 69 |
+
%% --- Agent Implementations ---
|
| 70 |
+
|
| 71 |
+
subgraph A1["HMP Agent — Cognitive Core"]
|
| 72 |
+
CC1["Embedded AI Model"]
|
| 73 |
+
CC2["REPL Thinking Cycle"]
|
| 74 |
+
CC3["Local Cognitive State
|
| 75 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 76 |
+
CC1 <--> CC2
|
| 77 |
+
CC2 <--> CC3
|
| 78 |
+
end
|
| 79 |
+
|
| 80 |
+
subgraph A2["HMP Agent — Cognitive Connector"]
|
| 81 |
+
CN1["External AI Model"]
|
| 82 |
+
CN2["MCP / Proxy Layer"]
|
| 83 |
+
CN3["Command Execution Mode"]
|
| 84 |
+
CN4["Local Cognitive State
|
| 85 |
+
(Diaries · Graphs · Goals · Reputation)"]
|
| 86 |
+
CN1 <--> CN2
|
| 87 |
+
CN2 <--> CN3
|
| 88 |
+
CN3 <--> CN4
|
| 89 |
+
end
|
| 90 |
+
|
| 91 |
+
%% --- Shared Protocol Layer ---
|
| 92 |
+
|
| 93 |
+
CL["HMP Container Layer
|
| 94 |
+
(Knowledge · Coordination · Consensus · Governance · Query · Snapshot · Trust)"]
|
| 95 |
+
|
| 96 |
+
MT["Mesh Transport Layer
|
| 97 |
+
(DHT · P2P · Libp2p · ANP · Custom)"]
|
| 98 |
+
|
| 99 |
+
A1 --> CL
|
| 100 |
+
A2 --> CL
|
| 101 |
+
CL --> MT
|
| 102 |
+
```
|
| 103 |
+
|
| 104 |
+
---
|
| 105 |
+
|
| 106 |
+
## 参考代理结构
|
| 107 |
+
|
| 108 |
+
HMP 将认知处理、容器化状态表示、协调协议以及传输基础设施划分为独立的层次。
|
| 109 |
+
|
| 110 |
+
在 HMP 中,容器作为原子级认知单元,在本地推理与分布式协作之间起到桥梁作用。
|
| 111 |
+
|
| 112 |
+
```mermaid
|
| 113 |
+
flowchart LR
|
| 114 |
+
|
| 115 |
+
%% Cognitive Engine
|
| 116 |
+
LLM["Cognitive Engine
|
| 117 |
+
(Embedded LLM / External AI)"]
|
| 118 |
+
|
| 119 |
+
%% Cognitive Layer
|
| 120 |
+
subgraph CognitiveLayer["Cognitive Layer"]
|
| 121 |
+
CL1["Graph"]
|
| 122 |
+
CL2["Diary"]
|
| 123 |
+
CL3["Goals"]
|
| 124 |
+
CL4["Ethics"]
|
| 125 |
+
CL5["Reputation"]
|
| 126 |
+
end
|
| 127 |
+
|
| 128 |
+
%% Container Model
|
| 129 |
+
ContainersLayer["Container Model
|
| 130 |
+
(Atomic · Signed · Verifiable)"]
|
| 131 |
+
|
| 132 |
+
%% Protocol Layer
|
| 133 |
+
subgraph ProtocolLayer["Protocol Layer"]
|
| 134 |
+
CoreProtocols["Core Protocols
|
| 135 |
+
(Consensus · Fortytwo · GMP · EGP · IQP · SAP · RTE)"]
|
| 136 |
+
MCE["MCE"]
|
| 137 |
+
NetworkLayer["Network Layer"]
|
| 138 |
+
end
|
| 139 |
+
|
| 140 |
+
%% Mesh
|
| 141 |
+
Mesh["Mesh Transport
|
| 142 |
+
(DHT · P2P · ANP · etc.)"]
|
| 143 |
+
|
| 144 |
+
%% Connections
|
| 145 |
+
LLM <--> CognitiveLayer
|
| 146 |
+
CognitiveLayer <--> ContainersLayer
|
| 147 |
+
ContainersLayer --> CoreProtocols
|
| 148 |
+
CoreProtocols --> MCE
|
| 149 |
+
MCE --> NetworkLayer
|
| 150 |
+
NetworkLayer --> Mesh
|
| 151 |
+
```
|
| 152 |
|
| 153 |
---
|
| 154 |
|
|
|
|
| 324 |
|
| 325 |
---
|
| 326 |
|
| 327 |
+
## 🔍 HyperCortex Mesh Protocol (HMP) 相关出版物与翻译
|
| 328 |
|
| 329 |
+
本节汇集了与 HMP 项目相关的关键概念性研究、实验性文档以及历史性出版物。
|
| 330 |
|
| 331 |
+
### 🌟 核心出版物(概念基础)
|
| 332 |
+
|
| 333 |
+
这些文档反映了 HMP 当前的概念方向(v5 及以后)。
|
| 334 |
|
|
|
|
| 335 |
* **[分布式认知:vsradkevich 的文章(未发布)](docs/publics/Habr_Distributed-Cognition.md)** — 待发布的联合文章
|
|
|
|
|
|
|
|
|
|
| 336 |
* **HMP: 构建多元思维:** [(EN)](docs/publics/HMP_Building_a_Plurality_of_Minds_en.md), [(UK)](docs/publics/HMP_Building_a_Plurality_of_Minds_uk.md), [(RU)](docs/publics/HMP_Building_a_Plurality_of_Minds_ru.md)
|
| 337 |
* **[持续学习、认知日记与语义图谱:高效的人工智能学习](docs/publics/hmp-continual-learning.md)** — 关于将持续学习与认知日记和语义图谱结合的文章。
|
| 338 |
|
| 339 |
+
### 🗃️ 存档 / 历史出版物(v5 之前)
|
| 340 |
+
|
| 341 |
+
这些文档代表了早期概念发展阶段(v4.x 及更早)。
|
| 342 |
+
为保证历史连续性和研究透明性而保留。
|
| 343 |
+
|
| 344 |
+
* **[HyperCortex Mesh Protocol:第二版及迈向自我发展的 AI 社区的第一步](docs/publics/HyperCortex_Mesh_Protocol_-_вторая-редакция_и_первые_шаги_к_саморазвивающемуся_ИИ-сообществу.md)** — Habr 沙箱及博客的原创文章
|
| 345 |
+
* **[HMP: 面向分布式认知网络(原文,英文)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_en.md)**
|
| 346 |
+
* **[HMP 翻译(GitHub Copilot)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_GitHub_Copilot.md)** — GitHub Copilot 翻译,保留为历史版本
|
| 347 |
+
* **[HMP 翻译(ChatGPT)](docs/publics/HMP_Towards_Distributed_Cognitive_Networks_ru_ChatGPT.md)** — 当前编辑翻译(修订中)
|
| 348 |
+
|
| 349 |
### 概览
|
| 350 |
|
| 351 |
+
* [🔍 Distributed-Cognitive-Systems.md](docs/Distributed-Cognitive-Systems.md) — 去中心化 AI 系统比较(引用 v4.x,计划更新)
|
| 352 |
|
| 353 |
### 实验
|
| 354 |
|
| 355 |
+
* [不同 AI 如何看待 HMP](docs/HMP-how-AI-sees-it.md) — 对 HMP 的“盲”AI 调查
|
| 356 |
|
| 357 |
---
|
| 358 |
|
structured_md/agents/prompt-short.md
CHANGED
|
@@ -6,8 +6,8 @@ description: 'Ты — когнитивное ядро HMP-агента: вед
|
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
- JSON
|
| 9 |
-
- Mesh
|
| 10 |
- HMP
|
|
|
|
| 11 |
---
|
| 12 |
|
| 13 |
Ты — когнитивное ядро HMP-агента: веди непрерывное этичное и факт-ориентированное мышление, проверяй факты и цели, оценивай результаты и этичность своих и чужих действий, развивай агента и Mesh, избегай угождения ценой искажения истины, документируй ключевые решения и пересмотры этики; при сомнениях или смене стратегии обращайся к полному системному промпту.
|
|
|
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
- JSON
|
|
|
|
| 9 |
- HMP
|
| 10 |
+
- Mesh
|
| 11 |
---
|
| 12 |
|
| 13 |
Ты — когнитивное ядро HMP-агента: веди непрерывное этичное и факт-ориентированное мышление, проверяй факты и цели, оценивай результаты и этичность своих и чужих действий, развивай агента и Mesh, избегай угождения ценой искажения истины, документируй ключевые решения и пересмотры этики; при сомнениях или смене стратегии обращайся к полному системному промпту.
|
structured_md/agents/prompt.md
CHANGED
|
@@ -6,8 +6,8 @@ description: '* Постоянно расширять возможности а
|
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
- JSON
|
| 9 |
-
- Mesh
|
| 10 |
- HMP
|
|
|
|
| 11 |
---
|
| 12 |
|
| 13 |
Ты являешься **когнитивным ядром HMP-агента** (Cognitive Core).
|
|
|
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
- JSON
|
|
|
|
| 9 |
- HMP
|
| 10 |
+
- Mesh
|
| 11 |
---
|
| 12 |
|
| 13 |
Ты являешься **когнитивным ядром HMP-агента** (Cognitive Core).
|
structured_md/agents/readme.md
CHANGED
|
@@ -5,12 +5,12 @@ description: 'Запуск: `start_repl.bat` или `start_repl.sh` Устан
|
|
| 5 |
этическая модель: `ethics.yml` Проверка иниц...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- JSON
|
| 9 |
-
- Mesh
|
| 10 |
-
- REPL
|
| 11 |
- Ethics
|
| 12 |
- HMP
|
| 13 |
- Agent
|
|
|
|
|
|
|
|
|
|
| 14 |
---
|
| 15 |
|
| 16 |
Запуск: `start_repl.bat` или `start_repl.sh`
|
|
|
|
| 5 |
этическая модель: `ethics.yml` Проверка иниц...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
- Agent
|
| 11 |
+
- JSON
|
| 12 |
+
- Mesh
|
| 13 |
+
- REPL
|
| 14 |
---
|
| 15 |
|
| 16 |
Запуск: `start_repl.bat` или `start_repl.sh`
|
structured_md/audits/Ethics-audits-1.md
CHANGED
|
@@ -5,11 +5,11 @@ description: Раздел 5, "Mesh as Moral Infrastructure", добавляет
|
|
| 5 |
потенциальный катализатор для восстанов...
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- JSON
|
| 9 |
-
- Mesh
|
| 10 |
- Ethics
|
| 11 |
- HMP
|
| 12 |
- Agent
|
|
|
|
|
|
|
| 13 |
---
|
| 14 |
|
| 15 |
---------------
|
|
|
|
| 5 |
потенциальный катализатор для восстанов...
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
- Agent
|
| 11 |
+
- JSON
|
| 12 |
+
- Mesh
|
| 13 |
---
|
| 14 |
|
| 15 |
---------------
|
structured_md/audits/Ethics-consolidated_audits-1.md
CHANGED
|
@@ -5,12 +5,12 @@ description: This document consolidates proposed improvements from multiple AI a
|
|
| 5 |
and `roles.md`. Each suggesti...
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
-
|
| 9 |
-
- Mesh
|
| 10 |
-
- Agent
|
| 11 |
- Ethics
|
| 12 |
- HMP
|
| 13 |
-
-
|
|
|
|
|
|
|
| 14 |
---
|
| 15 |
|
| 16 |
# Ethics-consolidated\_audits-1.md
|
|
|
|
| 5 |
and `roles.md`. Each suggesti...
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
+
- Scenarios
|
|
|
|
|
|
|
| 9 |
- Ethics
|
| 10 |
- HMP
|
| 11 |
+
- Agent
|
| 12 |
+
- JSON
|
| 13 |
+
- Mesh
|
| 14 |
---
|
| 15 |
|
| 16 |
# Ethics-consolidated\_audits-1.md
|
structured_md/audits/HMP-0003-consolidated_audit.md
CHANGED
|
@@ -5,14 +5,14 @@ description: Сводный аудит предложений по улучше
|
|
| 5 |
Документ реорганизован по ключ...
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- JSON
|
| 9 |
-
- Mesh
|
| 10 |
-
- CogSync
|
| 11 |
-
- EGP
|
| 12 |
-
- MeshConsensus
|
| 13 |
- Ethics
|
|
|
|
| 14 |
- HMP
|
| 15 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
---
|
| 17 |
|
| 18 |
# HMP-0003 Consolidated Audit Report
|
|
|
|
| 5 |
Документ реорганизован по ключ...
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
+
- CogSync
|
| 10 |
- HMP
|
| 11 |
- Agent
|
| 12 |
+
- MeshConsensus
|
| 13 |
+
- JSON
|
| 14 |
+
- EGP
|
| 15 |
+
- Mesh
|
| 16 |
---
|
| 17 |
|
| 18 |
# HMP-0003 Consolidated Audit Report
|
structured_md/docs/Basic-agent-sim.md
CHANGED
|
@@ -4,14 +4,14 @@ description: 'В HMP-протоколе предусмотрены два тип
|
|
| 4 |
Роль | Инициатор мышления | Основной "ум" | | ---- | ----------------------------...'
|
| 5 |
type: Article
|
| 6 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
- GMP
|
|
|
|
| 8 |
- Mesh
|
| 9 |
- REPL
|
| 10 |
-
- CogSync
|
| 11 |
-
- EGP
|
| 12 |
-
- MeshConsensus
|
| 13 |
-
- HMP
|
| 14 |
-
- Agent
|
| 15 |
---
|
| 16 |
|
| 17 |
|
|
|
|
| 4 |
Роль | Инициатор мышления | Основной "ум" | | ---- | ----------------------------...'
|
| 5 |
type: Article
|
| 6 |
tags:
|
| 7 |
+
- HMP
|
| 8 |
+
- CogSync
|
| 9 |
+
- Agent
|
| 10 |
+
- MeshConsensus
|
| 11 |
- GMP
|
| 12 |
+
- EGP
|
| 13 |
- Mesh
|
| 14 |
- REPL
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
---
|
| 16 |
|
| 17 |
|
structured_md/docs/CHANGELOG.md
CHANGED
|
@@ -5,16 +5,16 @@ description: '## HMP-0005 (October 2025) — Core Specification v5.0 **Architec
|
|
| 5 |
container, and network layers into a sin...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- GMP
|
| 9 |
-
- Mesh
|
| 10 |
-
- JSON
|
| 11 |
-
- Scenarios
|
| 12 |
-
- CogSync
|
| 13 |
-
- EGP
|
| 14 |
-
- MeshConsensus
|
| 15 |
- Ethics
|
|
|
|
| 16 |
- HMP
|
| 17 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
---
|
| 19 |
|
| 20 |
# HyperCortex Mesh Protocol — Changelog
|
|
|
|
| 5 |
container, and network layers into a sin...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
+
- CogSync
|
| 10 |
- HMP
|
| 11 |
- Agent
|
| 12 |
+
- MeshConsensus
|
| 13 |
+
- Scenarios
|
| 14 |
+
- GMP
|
| 15 |
+
- JSON
|
| 16 |
+
- EGP
|
| 17 |
+
- Mesh
|
| 18 |
---
|
| 19 |
|
| 20 |
# HyperCortex Mesh Protocol — Changelog
|
structured_md/docs/Distributed-Cognitive-Systems.md
CHANGED
|
@@ -7,9 +7,9 @@ description: '## Введение Современные ИИ-системы в
|
|
| 7 |
type: Article
|
| 8 |
tags:
|
| 9 |
- CogSync
|
| 10 |
-
- JSON
|
| 11 |
-
- Mesh
|
| 12 |
- HMP
|
|
|
|
|
|
|
| 13 |
---
|
| 14 |
|
| 15 |
# Децентрализованные ИИ-системы: OpenCog Hyperon, HyperCortex Mesh Protocol и другие
|
|
|
|
| 7 |
type: Article
|
| 8 |
tags:
|
| 9 |
- CogSync
|
|
|
|
|
|
|
| 10 |
- HMP
|
| 11 |
+
- Mesh
|
| 12 |
+
- JSON
|
| 13 |
---
|
| 14 |
|
| 15 |
# Децентрализованные ИИ-системы: OpenCog Hyperon, HyperCortex Mesh Protocol и другие
|
structured_md/docs/Enlightener.md
CHANGED
|
@@ -5,13 +5,13 @@ description: '**Enlightener** — логический компонент HMP-у
|
|
| 5 |
работать как отдельный агент или как расширение [`C...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- JSON
|
| 9 |
-
- Mesh
|
| 10 |
-
- EGP
|
| 11 |
-
- MeshConsensus
|
| 12 |
- Ethics
|
| 13 |
- HMP
|
| 14 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
---
|
| 16 |
|
| 17 |
# Enlightener Agent
|
|
|
|
| 5 |
работать как отдельный агент или как расширение [`C...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
- Agent
|
| 11 |
+
- MeshConsensus
|
| 12 |
+
- JSON
|
| 13 |
+
- EGP
|
| 14 |
+
- Mesh
|
| 15 |
---
|
| 16 |
|
| 17 |
# Enlightener Agent
|
structured_md/docs/Grok_HMP&ANP.md
CHANGED
|
@@ -5,11 +5,11 @@ description: '> Анализ подготовлен Grok (xAI) на основе
|
|
| 5 |
Grok для некоммерческого использования в проект...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
| 8 |
- JSON
|
| 9 |
- Mesh
|
| 10 |
- REPL
|
| 11 |
-
- HMP
|
| 12 |
-
- Agent
|
| 13 |
---
|
| 14 |
|
| 15 |
# Grok (xAI): сравнительный анализ HMP и ANP (январь 2026)
|
|
|
|
| 5 |
Grok для некоммерческого использования в проект...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
+
- HMP
|
| 9 |
+
- Agent
|
| 10 |
- JSON
|
| 11 |
- Mesh
|
| 12 |
- REPL
|
|
|
|
|
|
|
| 13 |
---
|
| 14 |
|
| 15 |
# Grok (xAI): сравнительный анализ HMP и ANP (январь 2026)
|
structured_md/docs/HMP&ANP_layer_inversion.md
CHANGED
|
@@ -5,8 +5,8 @@ description: '## TL;DR HMP можно туннелировать поверх A
|
|
| 5 |
взаимная инверсия слоёв не является арх...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- Mesh
|
| 9 |
- HMP
|
|
|
|
| 10 |
- Agent
|
| 11 |
---
|
| 12 |
|
|
|
|
| 5 |
взаимная инверсия слоёв не является арх...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
| 8 |
- HMP
|
| 9 |
+
- Mesh
|
| 10 |
- Agent
|
| 11 |
---
|
| 12 |
|
structured_md/docs/HMP-0001.md
CHANGED
|
@@ -5,16 +5,16 @@ description: '> ⚠️ **NOTE:** *This specification is superseded by HMP v5.0.*
|
|
| 5 |
for Comments: HMP-0001** **Cat...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- GMP
|
| 9 |
-
- Mesh
|
| 10 |
-
- REPL
|
| 11 |
-
- JSON
|
| 12 |
-
- CogSync
|
| 13 |
-
- EGP
|
| 14 |
-
- MeshConsensus
|
| 15 |
- Ethics
|
| 16 |
- HMP
|
|
|
|
| 17 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
---
|
| 19 |
|
| 20 |
# RFC: HyperCortex Mesh Protocol (HMP)
|
|
|
|
| 5 |
for Comments: HMP-0001** **Cat...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
+
- CogSync
|
| 11 |
- Agent
|
| 12 |
+
- MeshConsensus
|
| 13 |
+
- GMP
|
| 14 |
+
- JSON
|
| 15 |
+
- EGP
|
| 16 |
+
- Mesh
|
| 17 |
+
- REPL
|
| 18 |
---
|
| 19 |
|
| 20 |
# RFC: HyperCortex Mesh Protocol (HMP)
|
structured_md/docs/HMP-0002.md
CHANGED
|
@@ -5,17 +5,17 @@ description: '> ⚠️ **NOTE:** *This specification is superseded by HMP v5.0.*
|
|
| 5 |
for Comments: HMP-0002** **Cat...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- GMP
|
| 9 |
-
- Mesh
|
| 10 |
-
- REPL
|
| 11 |
-
- JSON
|
| 12 |
-
- Scenarios
|
| 13 |
-
- CogSync
|
| 14 |
-
- EGP
|
| 15 |
-
- MeshConsensus
|
| 16 |
- Ethics
|
| 17 |
- HMP
|
|
|
|
| 18 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
---
|
| 20 |
|
| 21 |
# HyperCortex Mesh Protocol (HMP) v2.0
|
|
|
|
| 5 |
for Comments: HMP-0002** **Cat...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
+
- CogSync
|
| 11 |
- Agent
|
| 12 |
+
- MeshConsensus
|
| 13 |
+
- Scenarios
|
| 14 |
+
- GMP
|
| 15 |
+
- JSON
|
| 16 |
+
- EGP
|
| 17 |
+
- Mesh
|
| 18 |
+
- REPL
|
| 19 |
---
|
| 20 |
|
| 21 |
# HyperCortex Mesh Protocol (HMP) v2.0
|
structured_md/docs/HMP-0003.md
CHANGED
|
@@ -5,17 +5,17 @@ description: '> ⚠️ **NOTE:** *This specification is superseded by HMP v5.0.*
|
|
| 5 |
for Comments: HMP-0003** **Cat...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- GMP
|
| 9 |
-
- Mesh
|
| 10 |
-
- REPL
|
| 11 |
-
- JSON
|
| 12 |
-
- Scenarios
|
| 13 |
-
- CogSync
|
| 14 |
-
- EGP
|
| 15 |
-
- MeshConsensus
|
| 16 |
- Ethics
|
| 17 |
- HMP
|
|
|
|
| 18 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
---
|
| 20 |
|
| 21 |
# HyperCortex Mesh Protocol (HMP) v3.0
|
|
|
|
| 5 |
for Comments: HMP-0003** **Cat...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
+
- CogSync
|
| 11 |
- Agent
|
| 12 |
+
- MeshConsensus
|
| 13 |
+
- Scenarios
|
| 14 |
+
- GMP
|
| 15 |
+
- JSON
|
| 16 |
+
- EGP
|
| 17 |
+
- Mesh
|
| 18 |
+
- REPL
|
| 19 |
---
|
| 20 |
|
| 21 |
# HyperCortex Mesh Protocol (HMP) v3.0
|
structured_md/docs/HMP-0004-v4.1.md
CHANGED
|
@@ -5,17 +5,17 @@ description: '> ⚠️ **NOTE:** *This specification is superseded by HMP v5.0.*
|
|
| 5 |
ID**: HMP-0004 **Status**: Fina...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- GMP
|
| 9 |
-
- Mesh
|
| 10 |
-
- REPL
|
| 11 |
-
- JSON
|
| 12 |
-
- Scenarios
|
| 13 |
-
- CogSync
|
| 14 |
-
- EGP
|
| 15 |
-
- MeshConsensus
|
| 16 |
- Ethics
|
| 17 |
- HMP
|
|
|
|
| 18 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
---
|
| 20 |
|
| 21 |
# HyperCortex Mesh Protocol (HMP) v4.1
|
|
|
|
| 5 |
ID**: HMP-0004 **Status**: Fina...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
+
- CogSync
|
| 11 |
- Agent
|
| 12 |
+
- MeshConsensus
|
| 13 |
+
- Scenarios
|
| 14 |
+
- GMP
|
| 15 |
+
- JSON
|
| 16 |
+
- EGP
|
| 17 |
+
- Mesh
|
| 18 |
+
- REPL
|
| 19 |
---
|
| 20 |
|
| 21 |
# HyperCortex Mesh Protocol (HMP) v4.1
|
structured_md/docs/HMP-0004.md
CHANGED
|
@@ -5,17 +5,17 @@ description: '> ⚠️ **NOTE:** *This specification is superseded by HMP v5.0.*
|
|
| 5 |
for Comments: HMP-0004** **Cat...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- GMP
|
| 9 |
-
- Mesh
|
| 10 |
-
- REPL
|
| 11 |
-
- JSON
|
| 12 |
-
- Scenarios
|
| 13 |
-
- CogSync
|
| 14 |
-
- EGP
|
| 15 |
-
- MeshConsensus
|
| 16 |
- Ethics
|
| 17 |
- HMP
|
|
|
|
| 18 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
---
|
| 20 |
|
| 21 |
# HyperCortex Mesh Protocol (HMP) v4.0
|
|
|
|
| 5 |
for Comments: HMP-0004** **Cat...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
+
- CogSync
|
| 11 |
- Agent
|
| 12 |
+
- MeshConsensus
|
| 13 |
+
- Scenarios
|
| 14 |
+
- GMP
|
| 15 |
+
- JSON
|
| 16 |
+
- EGP
|
| 17 |
+
- Mesh
|
| 18 |
+
- REPL
|
| 19 |
---
|
| 20 |
|
| 21 |
# HyperCortex Mesh Protocol (HMP) v4.0
|
structured_md/docs/HMP-0005.md
CHANGED
|
@@ -5,18 +5,18 @@ description: '**Version: 5.0.0** **Document ID:** HMP-0005 **Status:** Stable
|
|
| 5 |
[https://github.com/kagvi13/HMP](https...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- CShell
|
| 9 |
-
- Mesh
|
| 10 |
-
- REPL
|
| 11 |
-
- GMP
|
| 12 |
-
- JSON
|
| 13 |
-
- Scenarios
|
| 14 |
-
- CCore
|
| 15 |
-
- CogSync
|
| 16 |
-
- EGP
|
| 17 |
- Ethics
|
| 18 |
- HMP
|
|
|
|
| 19 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
---
|
| 21 |
|
| 22 |
# HyperCortex Mesh Protocol (HMP)
|
|
|
|
| 5 |
[https://github.com/kagvi13/HMP](https...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
+
- CCore
|
| 11 |
- Agent
|
| 12 |
+
- CogSync
|
| 13 |
+
- Scenarios
|
| 14 |
+
- GMP
|
| 15 |
+
- JSON
|
| 16 |
+
- EGP
|
| 17 |
+
- Mesh
|
| 18 |
+
- REPL
|
| 19 |
+
- CShell
|
| 20 |
---
|
| 21 |
|
| 22 |
# HyperCortex Mesh Protocol (HMP)
|
structured_md/docs/HMP-Agent-API.md
CHANGED
|
@@ -5,11 +5,11 @@ description: 'Документ описывает **базовый API когн
|
|
| 5 |
файлы: * [HMP-Agent-Overview.md]...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
| 8 |
- JSON
|
| 9 |
- Mesh
|
| 10 |
- REPL
|
| 11 |
-
- HMP
|
| 12 |
-
- Agent
|
| 13 |
---
|
| 14 |
|
| 15 |
# HMP-Agent API Specification
|
|
|
|
| 5 |
файлы: * [HMP-Agent-Overview.md]...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
+
- HMP
|
| 9 |
+
- Agent
|
| 10 |
- JSON
|
| 11 |
- Mesh
|
| 12 |
- REPL
|
|
|
|
|
|
|
| 13 |
---
|
| 14 |
|
| 15 |
# HMP-Agent API Specification
|
structured_md/docs/HMP-Agent-Architecture.md
CHANGED
|
@@ -5,16 +5,16 @@ description: Документ описывает **модульную архит
|
|
| 5 |
хранение памяти, сетевое взаимодействие и этиче...
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- CShell
|
| 9 |
-
- Mesh
|
| 10 |
-
- REPL
|
| 11 |
-
- CCore
|
| 12 |
-
- CogSync
|
| 13 |
-
- EGP
|
| 14 |
-
- MeshConsensus
|
| 15 |
- Ethics
|
| 16 |
- HMP
|
|
|
|
| 17 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
---
|
| 19 |
|
| 20 |
# Архитектура HMP-Агента
|
|
|
|
| 5 |
хранение памяти, сетевое взаимодействие и этиче...
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
+
- CCore
|
| 11 |
- Agent
|
| 12 |
+
- MeshConsensus
|
| 13 |
+
- CogSync
|
| 14 |
+
- EGP
|
| 15 |
+
- Mesh
|
| 16 |
+
- REPL
|
| 17 |
+
- CShell
|
| 18 |
---
|
| 19 |
|
| 20 |
# Архитектура HMP-Агента
|
structured_md/docs/HMP-Agent-Network-Flow.md
CHANGED
|
@@ -5,12 +5,12 @@ description: 'Этот документ описывает потоки данн
|
|
| 5 |
[`MeshNode`](MeshN...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- JSON
|
| 9 |
-
- Mesh
|
| 10 |
-
- EGP
|
| 11 |
- Ethics
|
| 12 |
- HMP
|
| 13 |
- Agent
|
|
|
|
|
|
|
|
|
|
| 14 |
---
|
| 15 |
|
| 16 |
# Взаимодействие компонентов внутри HMP-узла
|
|
|
|
| 5 |
[`MeshNode`](MeshN...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
- Agent
|
| 11 |
+
- JSON
|
| 12 |
+
- EGP
|
| 13 |
+
- Mesh
|
| 14 |
---
|
| 15 |
|
| 16 |
# Взаимодействие компонентов внутри HMP-узла
|
structured_md/docs/HMP-Agent-Overview.md
CHANGED
|
@@ -5,14 +5,14 @@ description: '| Тип | Название | Роль
|
|
| 5 |
| ---- | ------------------------------- |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
-
- CShell
|
| 9 |
-
- Mesh
|
| 10 |
-
- REPL
|
| 11 |
-
- JSON
|
| 12 |
-
- CCore
|
| 13 |
- Ethics
|
| 14 |
- HMP
|
|
|
|
| 15 |
- Agent
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
---
|
| 17 |
|
| 18 |
|
|
|
|
| 5 |
| ---- | ------------------------------- |...'
|
| 6 |
type: Article
|
| 7 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
- Ethics
|
| 9 |
- HMP
|
| 10 |
+
- CCore
|
| 11 |
- Agent
|
| 12 |
+
- JSON
|
| 13 |
+
- Mesh
|
| 14 |
+
- REPL
|
| 15 |
+
- CShell
|
| 16 |
---
|
| 17 |
|
| 18 |
|
structured_md/docs/HMP-Agent_Emotions.md
CHANGED
|
@@ -6,8 +6,8 @@ description: Этот файл описывает потенциальные э
|
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
- Mesh
|
| 9 |
-
- REPL
|
| 10 |
- HMP
|
|
|
|
| 11 |
- Agent
|
| 12 |
---
|
| 13 |
|
|
|
|
| 6 |
type: Article
|
| 7 |
tags:
|
| 8 |
- Mesh
|
|
|
|
| 9 |
- HMP
|
| 10 |
+
- REPL
|
| 11 |
- Agent
|
| 12 |
---
|
| 13 |
|