KubeVela

An application-centric delivery control plane that turns one OAM Application into multi-cluster Kubernetes resources through composable CUE modules.

• Category: App Definition & GitOps
• CNCF maturity: Incubating
• Language: Go
• License: Apache-2.0
• Repository: kubevela/kubevela
• Documented at commit: a10dba6 (master, 2026-06-10)

What it is

KubeVela is a Kubernetes controller built on the Open Application Model (OAM). A user writes one Application custom resource that lists components, policies, and a workflow, and the controller expands it into the real Kubernetes objects that run the workload. The abstraction layer is written in CUE templates rather than Go, so a platform team can add new component and trait types by adding CUE definitions instead of rebuilding the controller.

OAM splits an application into Components, which model a workload, and Traits, which attach operational capabilities to a component (source: CNCF blog 2023-03-31). KubeVela was the first implementation of OAM, proposed in 2020 by Alibaba Cloud and Microsoft Azure. It sits above raw Kubernetes manifests and above Helm: it can deliver a Helm chart, a Terraform module, or a plain workload, then track and garbage-collect everything it applied.

The project ships three entry points: a controller-manager, a vela CLI, and a kubectl plugin. The controller is the part that reconciles Application resources; the CLI is for operators who drive delivery from a terminal.

When to use it

• You are building an internal developer platform and want one self-service application API instead of exposing raw Deployments, Services, and Ingresses.
• You need to deliver the same application across multiple clusters with placement policies from a single control plane.
• You want platform engineers to define reusable component and trait abstractions in CUE without forking or rebuilding a controller.
• You already standardize on Helm charts or Terraform modules and want to wrap them in a higher-level, workflow-driven delivery model.

When it is not the right fit: if you only need to sync raw manifests from Git, a plain GitOps syncer is simpler. If your team has no appetite for CUE, the abstraction layer adds learning cost that lands on the platform side.

In this deep-dive

• History: origin, milestones, and why it exists.
• Architecture: components and how requests flow.
• Adoption & Ecosystem: who runs it and what surrounds it.
• Internals: the code paths that matter, read from source.
• Getting Started: install and a first working setup.

Sources

1. kubevela/kubevela repository, commit a10dba6
2. kubevela/community ADOPTERS.md
3. kubevela/community GOVERNANCE.md
4. CNCF: KubeVela brings software delivery control plane capabilities to CNCF Incubator (2023-02-27)
5. CNCF: KubeVela, the road to cloud native application and platform engineering (2023-03-31)
6. CNCF project page: KubeVela
7. KubeVela Documentation: Introduction