--:--:--
Servers Online
5/5
All nodes healthy
Avg CPU
34%
Across fleet
Avg RAM
61%
Across fleet
Memory Parts
7
Active AI modules
Services
24
Running containers
Functions
18
Deployed endpoints
Webhooks
142
Fired today
Alerts
2
Require attention
AI Automation Pipeline
Webhook In
N8N Router
OpenClaw
Claude API
Gemini API
Vector DB
Agent Pool
Webhook Out
Context / Pack Evidence
Fleet Resource Usage
Live Log Stream
AI Brain Memory Map — Nodes & Linking

Deployed Functions & Code Modules
Running Services & Containers
Service Server Port Type Status CPU RAM Uptime
Fleet Wiring & Connection Map

Drag nodes · Scroll to zoom · Hover for details · Live connection topology

Connection Legend
API / REST Webhook SSH Tunnel Data Stream Monitoring LLM API
AI Memory Architecture Diagram

Full memory hierarchy — layers, data flows, and links between all memory subsystems

Write Path

Agent output → STM buffer → Embedding (OpenAI/local model) → Vector Store → LTM index. Episodic events also written with timestamp + context tag.

Recall Path

Query → Embed → Cosine similarity search in Vector Store → top-K chunks retrieved from LTM + Episodic → injected into Core Brain prompt context.

Sync Cycle

Every 5 min: STM eviction of stale entries. Daily: LTM consolidation pass. N8N triggers semantic indexing. Crypto state written on every price tick event.