Getting Started
Quick start guide for cluster requirements, installation, and first agent.
Cluster Requirements
Kaigents runs on any standard Kubernetes cluster. For optimal performance with AI workloads:
- Minimum 4 CPU cores
- 8GB RAM
- 100GB storage (SSD recommended)
- Kubernetes v1.25+
For sovereign AI compute on AMD Ryzen AI hardware, ensure your nodes have:
- AMD Ryzen AI processors
- Support for OpenWeight models
- 16GB+ RAM for large models
Installation
Install Kaigents using our Helm chart:
# Add the Kaigents Helm repository helm repo add kaigents https://charts.kaigents.dev helm repo update # Install Kaigents helm install kaigents kaigents/kaigents
For detailed installation instructions, including customization options, see our documentation.
First Agent Example
Create your first agent using a simple YAML definition:
apiVersion: kaigents.dev/v1
kind: Agent
metadata:
name: sample-agent
spec:
model: qwen3.6-35b
skills:
- name: documentation-skill
memory:
shortTerm: true
longTerm: trueApply the definition to your cluster:
kubectl apply -f sample-agent.yaml
Next Steps
- Explore our Skills system to define quality gates
- Learn about durable execution with Temporal
- Configure S3 artifact storage for outputs
- Set up hybrid execution across CPU, GPU, and NPU
Ready to dive deeper? Check out our full documentation for advanced configuration and deployment strategies.