and build a platform application teams love
to learn more about Kubestack
Kubestack captures proven architecture decisions for the platform you're building, and the Kubestack skill teaches your coding agent to scaffold the OpenTofu code for your entire stack within them. An agent working on raw OpenTofu quickly invents multiple different ways to do the same task, one per session. The framework's conventions give every agent session, and every engineer, one stable pattern to follow. Engineers new to OpenTofu or Kubernetes rely on the framework defaults to make progress now and learn as they go. Experienced platform builders use their agent to skip the repetitive boilerplate.
The cloud native landscape moves fast and so do your application teams. You need a way to continuously evolve your platform without sacrificing availability. Kubestack's separation of concerns is what makes agent-driven platform engineering safe: your agent only edits and validates files locally, while the GitOps pipeline owns every change to real infrastructure. Each change is peer reviewed alongside its OpenTofu plan, validated against the internal ops environment, and only then promoted to the environments running application workloads. Non-deterministic LLM behavior can produce a bad proposal, but it can never directly cause an incident.
Being part of an open-source community of platform builders gives you access to future improvements and lowers your maintenance effort. But ultimately your platform is defined by native OpenTofu code in your repository. Your agent, like your engineers, works on plain OpenTofu modules, not on a proprietary abstraction. Contribute upstream, extend your platform with custom OpenTofu code, replace individual Kubestack modules or even fork the entire open-source framework. You own the code, you are in control.
and build a platform application teams love
to learn more about Kubestack
Kubestack builds on each cloud's managed Kubernetes offering for the best possible integration and support, but unifies platform features, developer experience and runtime environments for your application developers across all clusters and clouds. Your agent uses the same conventions everywhere, so a multi-cloud platform is no harder to maintain than a single-cloud one.
Context-aware access automation for humans and AI agents


