CuteOS brings AI-assisted operations to real infrastructure without giving up control. Describe the outcome you want, review the plan, and approve what runs.
Behind the interface is a cloud-native platform with policy, secrets, audit, and MCP-connected integrations built for serious environments.
CuteOS turns plain-English requests into controlled infrastructure workflows across the systems teams already run.
Ask for the outcome you want and let CuteOS turn it into a structured, reviewable change plan.
Connect networking, virtualization, Kubernetes, DNS, and other systems through MCP APIs and supported integrations.
OPA policy checks, approval gates, and audit records keep execution controlled instead of letting AI run loose.
A Go-based control plane on Kubernetes coordinates workflows, services, jobs, and execution boundaries.
OpenBao handles secrets while NATS powers event-driven coordination across the platform.
Start in a safe lab, prove the workflow, then take the same platform into more controlled production environments.
CuteOS is not a chat demo for infrastructure. It is built for environments where approvals, policy, secrets, and audit actually matter.
Under the hood, a cloud-native control plane, OPA policy enforcement, OpenBao secrets, and event-driven workflows on NATS keep the platform usable in real operations.
The recommended first install path is Docker Desktop Kubernetes on macOS. The full install guide also covers Windows and an operator-managed Helm path for existing Kubernetes.
Public community install path for local lab and test environments.
curl -fsSL https://install.cuteos.io/install.sh | bashInstalls CuteOS onto a local Kubernetes cluster, starts local port-forwards by default, and prints the dashboard URL plus the admin password at the end.
Start with a safe local environment, build trust in the workflow, and bring the same operating model into larger teams, clusters, and production change processes.
Infrastructure AI