Genetic Agent Systems

Setup guide for genetic agent systems on ABUZ8 OS. From basic configuration to advanced orchestration.

EARLY ACCESS

ABUZ8 OS is in early access — there is no public download yet. Request access and you'll receive a build directly. The steps below apply once you have one.

🚀 Basic Setup

STEP 1

Install ABUZ8 OS

Early-access builds are delivered after your access request is approved. Run the installer or extract the archive you receive.

# Windows — run the installer you received
cd Downloads
.\ABUZ8_OS_setup.exe

# Or extract from 7z
7z x ABUZ8_OS.7z -oC:\ABUZ8
STEP 2

Launch Dashboard

Run the dashboard and verify all services are online.

cd C:\ABUZ8\dist\win-unpacked
.\ABUZ8.exe

# Verify at http://127.0.0.1:8900/api/dashboard
curl http://127.0.0.1:8900/api/dashboard
STEP 3

Configure GPU

Verify your NVIDIA GPU (or GPUs) are detected and operational. Local models need VRAM — the more, the larger the model you can run.

nvidia-smi

# Lists every detected NVIDIA GPU
# with its VRAM and driver version
STEP 4

Load Models

Select and load AI models for your agent systems.

# Via dashboard UI or API
curl -X POST http://127.0.0.1:8900/api/tools/run \
  -H "Content-Type: application/json" \
  -d '{"name":"load_model","args":{"model":"llama-3.1-70b"}}'

⚙️ Configuration

models.config.json

Define which models to load and their priorities.

{"primary": "llama-3.1-70b", "secondary": "gemma-2-27b"}

agents.config.json

Configure agent behavior and orchestration rules.

{"mode": "agentic", "verify": true, "reflect": true}

gpu.config.json

GPU allocation and memory management settings.

{"allocation": "auto", "memory_limit": "20GB"}

network.config.json

Network settings for external integrations.

{"mcp": true, "api_keys": "encrypted"}

🔬 Advanced Configuration