Skip to main content

​
When should you use it

Use the Kubernetes MCP server when you want to:
  • Manage Kubernetes and OpenShift clusters through AI assistants.
  • Perform CRUD operations on Kubernetes resources (pods, deployments, services, etc.).
  • Work with Helm charts (install, list, uninstall).
  • Query logs, metrics, and events from the cluster.
  • Automate operational workflows like resource scaling, monitoring, and troubleshooting.
  • Extend support to OpenShift with projects and additional resource types.

​
Requirements

  • Supported platforms: Kubernetes and OpenShift.
  • Configuration:
    • Uses ~/.kube/config or in-cluster config.
    • Automatically detects changes and reloads.
  • Dependencies: Kubernetes cluster access with sufficient RBAC permissions.

​
Tools

​
Configuration

configuration_view — View the current Kubernetes configuration (full or minified).

​
Events

events_list — List Kubernetes events across namespaces (or filtered by namespace).

​
Helm

helm_install — Install a Helm chart. helm_list — List Helm releases (all or specific namespaces). helm_uninstall — Uninstall a Helm release.

​
Namespaces & Projects

namespaces_list — List all namespaces in the cluster. projects_list — List all OpenShift projects.

​
Pods

pods_list — List all pods in the cluster. pods_list_in_namespace — List pods in a specific namespace. pods_get — Get details of a specific pod. pods_log — Retrieve logs from a pod (with container and previous options). pods_exec — Execute commands inside a pod container. pods_delete — Delete a pod by name. pods_run — Run a new pod with a specified image. pods_top — Retrieve pod CPU/memory usage (via metrics server).

​
Resources (Generic CRUD)

resources_create_or_update — Create or update a Kubernetes resource from YAML/JSON. resources_get — Retrieve a specific resource. resources_list — List resources (by API version, kind, namespace, label selectors). resources_delete — Delete a resource by kind and name.

​
Dashboards & Monitoring

pods_top — Get resource usage for pods (cluster-wide or namespace-specific).

​
Notes

  • Works with both self-managed Kubernetes clusters and OpenShift.
  • Supports Helm operations natively for release management.
  • RBAC permissions must match the requested operations (e.g., creating pods, deleting resources).
  • Ideal for use cases where AI agents need to act as cluster copilots for devops, troubleshooting, and automation.

Portkey is now PRISMA AIRS AI Gateway. See it in action.

Contact Us
Last modified on August 3, 2026