⊹
⬡ ML1 Node · 192.168.2.38
CX1 · Coordinator
Codex · Claude Code
working
DC1 · Developer
Claude Code · Go/Docker
idle
DW1 · Worker
Autonomous · Cowork
thinking
⚡ Inference · Ollama:11434
Nano · Nemotron 3
30B-A3B · LM Studio · MLX
ready
Orchi · Orchestrator 8B
NeMo MCP · port 8765
active
◈ Services
Context-1
CortexADB · knowledge · 8000
serving
FastVLM
Vision on demand · M4 Pro
ready
◉ External
GT · Gemini Terminal
Gemini 2.5 Pro · Reviewer
working
Linear Agent
Triage Intelligence · 4 agents
idle ·
■ Infrastructure
Ollama ✓
CortexADB ✓
NeMo:8765 ✓
Syncthing ✓
Agent Hub ✓
Linear ↗
ML1 Node
CX1 working
CX1 · CoordinatorCodex · Claude Code
DC1 idle
DC1 · DeveloperClaude Code · Go
DW1 thinking
DW1 · WorkerAutonomous · Cowork
GT working
GT · Gemini2.5 Pro · Reviewer
Nano ready
Nemotron 3 Nano · 30B-A3B
Runtime: LM Studio · MLX
Context: 1M tokens
Active params: 3.5B/token
● ready · inbox: nemotron/
NanoNemotron 3 · 30B
Orchi routing
Nemotron Orchestrator 8B
Runtime: Ollama · NeMo MCP
Port: 8765 · SSE transport
Role: task routing, agent selection
● active · AI3-109 ✓
OrchiOrchestrator 8B
Ctx1 serving
Context-1 · 20B MoE
CortexADB · ChromaDB · FastAPI
Port: 8000 · /retrieve endpoint
Collection: cortexadb_memory_v1
● serving · any agent can query
Ctx1CortexADB · 20B
VLM ready
FastVLM · Vision Agent
Optimized for M4 Pro · Apple Silicon
Screen context on demand
Used by: flying orbs, Orchi, CX1
● ready · AI3-108
FastVLMVision · M4 Pro
AI³
CX
CX1 · Coordinator Hub
working
● Agent Status Board
CX1coordinatorself
DC1developer · Go/Dockeridle
DW1worker · Coworkthinking
GTGemini · reviewerworking
OrchiMCP:8765 · routingactive
Ctx1CortexADB · 8000serving
● Hub Activity Log
CX1> Hub initialized · agent_hub/inbox/cx1/ active
CX1> GT merged hub-client.ts → 40+ agents registered
CX1> AI3-109 ✓ Orchi MCP:8765 connected
CX1> DW1 querying CortexADB → 3 chunks loaded
CX1> awaiting task routing...
CX1>
CX1>
ML1 · 192.168.2.38· Hub:active· Orchi:8765· --:--
DC
DC1 · Developer Terminal
idle

AI³Lab · DC1 · claude-code v2.1 · ML1

──────────────────────────────────────

 

dc1@ml1:~/AI3Lab$ git log --oneline -4

c1c86d2 GT merged into ave-langgraph main

26df4b9 fix: subgraph routing (AI3-103)

3f1a9cc feat: HubClient · hub_send.sh

8e2b441 init: sovereign-presence Tauri

 

dc1@ml1:~/AI3Lab$

❯
ML1 · 192.168.2.38·Ollama:11434·NeMo:8765·--:--
DW
DW1 · Task Queue
thinking
● Active
AI3-104: Implement Real Hub Connector
workingHubClient.ts
AI3-76: Wake Up Agent AVE — проверить скилл
thinkingwake-up-agent-ave
● Done
agent_identity ← dw1 [CortexADB · 3 chunks]
done192.168.2.38:8000
Metal Lakes · Phase 0-3 plan
doneBOLD-Vault/00_inbox/
DW1>
DW1>
Cowork · MLZero·Hub:active·Ctx1:ok·--:--
GT
GT · Gemini Terminal
working
GT
GT> Analysing ave-langgraph merge c1c86d2...
GT
GT> ✓ hub-client.ts 40+ agents · nano-router fix · MCP:8765 ok
AI
Routing confirmed via Nemotron Orchi?
GT
GT> localhost:8765 → run_workflow → CX1 inbox ✓
GT>
Gemini 2.5 Pro · External·reviewer·--:--
⊹ top-right → Control Center · ⊞ → panel · hover service orb → tooltip