Getting Started with OpenKubes
Your Kubernetes Framework. Built for the Age of AI.
OpenKubes helps you build your own Kubernetes platform with open, modular components and explicit platform contracts. Choose the path that matches what you want to explore. You do not need a bare-metal cluster to start learning.
Choose your path
| Path | Best for | Start here |
|---|---|---|
| OpenKubes Local | Developers evaluating OpenKubes on a Mac | Run OpenKubes Local |
| OpenKubes Cluster | Platform engineers exploring cluster lifecycle and infrastructure choices | Explore OK-Cluster |
| AI & Agentic Operations | Teams running AI workloads and evaluating future agent-assisted operations | Explore AI capabilities |
1. OpenKubes Local
Recommended first step if you are new to OpenKubes.
OpenKubes Local provides a development environment with separate management and infrastructure/workload Kubernetes clusters on macOS using Multipass and K3s. The accompanying tutorials introduce KubeVirt, Cluster API and Crossplane.
Before starting, review the local prerequisites.
git clone https://github.com/openkubes/ok-local.git
cd ok-local
make setup
The setup requires the tools and host resources described in the repository. Once it finishes, use make help and the OK Local tutorials to continue with cluster access, KubeVirt, CAPI and Crossplane.
2. OpenKubes Cluster
For infrastructure and platform engineers.
OK-Cluster is the OpenKubes cluster-lifecycle engine. Its scope includes declarative cluster creation, observation and upgrades across supported infrastructure and OS paths. It is not a one-command replacement for the local tutorial: cluster creation requires explicit environment inputs, operator review and authorization.
Start with the OK-Cluster README, then consult the operator runbook before attempting any live cluster operation.
For an initial, offline, non-mutating exploration, the repository documents:
git clone https://github.com/openkubes/ok-cluster.git
cd ok-cluster
make ok-cli-build
.tools/bin/ok cluster create --help
This builds and inspects the CLI; it does not provision a cluster. For runtime requirements and supported workflows, follow the repository documentation rather than copying old bare-metal installation commands.
3. AI and Agentic Operations
For AI workloads, AI-enabled platform operations and future agent integrations.
OpenKubes distinguishes running AI workloads on Kubernetes from using AI agents to propose or coordinate infrastructure operations. Agentic Operations is an evolving, optional capability: policy checks, explicit authorization, reconciliation and evidence remain essential. Do not assume autonomous production operations are available as an out-of-the-box installation.
Start with the Agentic Operations Reference Architecture and the OpenKubes architecture overview.
We will link a dedicated AI installation guide here once its source repository and supported entry points are confirmed.
Where to go next
- Developer documentation — documentation collected from OpenKubes repositories.
- Features — capabilities of the framework.
- Architecture — design principles, platform contracts and boundaries.
- GitHub organization — repositories, issues and contributions.
Source of truth: This page lives in
openkubes/openkubes/docs/getting-started/README.md. Component-specific setup details remain in their own repositories.