Guide · Agent skill
How to use the GKE Cluster Creation agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The gke-cluster-creation skill teaches an AI agent how to plan and create a Google Kubernetes Engine (GKE) cluster, which is a managed group of servers that runs containerised apps on Google Cloud. This guide is for people who have not used agent skills before and want a working setup. By the end you will know how to install the skill, how to ask for a cluster in plain language, and what the agent will check and confirm with you before it creates anything.
Quick start
- 1Download and unzip the skill. Download the ZIP from this page and unzip it into ~/.claude/skills/ or your project's .claude/skills/ folder.
- 2Check your Google Cloud access. Confirm you have a project with GKE permissions and that the gcloud CLI is installed and authenticated.
- 3Ask for a cluster in plain language. Describe the cluster you want, including the project, region and cluster name, or let the agent ask you for them.
- 4Choose a mode. Ask the agent to explain Autopilot versus Standard and pick one. Autopilot is the default unless you need Standard's control.
- 5Review the default configuration. Read the proposed gcloud command or create_cluster payload and check each setting before confirming.
- 6Create and track. Let the agent create the cluster, then watch progress through the operation tools.
- 7Verify the settings. Ask the agent to check the cluster with get_cluster so you can confirm the golden path settings were applied.
What is an agent skill and what does this one do?
An agent skill is a folder of instructions and reference files that an AI agent loads when a task matches the skill's description. This one is designed for the job of creating GKE clusters, and it also covers production readiness audits of existing clusters.
Once installed, the agent walks through the setup. It asks for the Google Cloud project, the region or zone, and the cluster name. It explains the difference between Autopilot and Standard modes, and it proposes a pre-defined secure default configuration for you to confirm before anything is created.
- Autopilot is Google-managed node infrastructure, billed per pod resource request. It is the default the skill recommends for most production workloads.
- Standard gives you self-managed nodes, billed per node (VM), with full control over node configuration, kernel tuning and custom OS.
- The default configuration includes private nodes, release channel guidance, master authorized networks, and Workload Identity guidance, among other settings.
How do you install the gke-cluster-creation skill?
Download the skill ZIP from this page and unzip it. Place the resulting skill folder in one of two locations. Use ~/.claude/skills/ to make it available across your projects, or use .claude/skills/ inside a specific project folder to make it available only there.
If you use the Claude apps instead, open Settings, go to Capabilities, and upload the skill under Skills.
- Personal install: unzip into ~/.claude/skills/
- Project install: unzip into the .claude/skills/ folder of your project
- Claude apps: Settings → Capabilities → Skills, then upload the skill
What do you need before the skill can create a cluster?
You need a Google Cloud project where you have GKE permissions. Creating clusters costs money and needs IAM access, so check with your account owner before you start if you are unsure.
For the CLI commands the skill uses, the gcloud CLI must be installed and authenticated. The skill can also use optional MCP tools: list_clusters, create_cluster, get_cluster, list_operations and get_operation. The skill does not run scripts, but it does use the network to reach Google Cloud.
- A Google Cloud project with GKE permissions
- gcloud CLI installed and authenticated (needed for CLI commands)
- Optional MCP tools: list_clusters, create_cluster, get_cluster, list_operations, get_operation
How do you trigger the skill with a natural request?
You do not need special syntax. Describe what you want in plain language that mentions clusters, GKE, or cluster mode choices. The skill's description covers creating clusters, provisioning GKE environments, selecting cluster modes and auditing clusters.
If your request leaves out the project, region or cluster name, the agent should ask you for them before taking action. Do not expect it to guess these values.
- Set up a production GKE cluster in europe-west2 for my payments API in project my-project.
- Should I use GKE Autopilot or Standard for a workload that needs a custom kernel?
- Audit my existing GKE cluster against production best practices.
- Create a dev GKE cluster quickly for testing; it does not need hardening.
What does the agent do during a cluster creation?
The skill follows a fixed workflow. The agent first discovers context by checking existing clusters, then gathers inputs and selects a mode. It configures networking, using an auto-created subnet by default or one you bring. Before creating anything, it shows the default configuration block and waits for your confirmation.
After creation, the agent tracks progress with the operation tools and then verifies that the golden path settings were applied to the cluster.
- Discover context: check existing clusters
- Gather inputs: project, location, cluster name, environment type
- Select mode and explain trade-offs
- Configure networking
- Review settings and confirm
- Create, track, then verify
What does the golden path Autopilot command look like?
The skill's default template creates an Autopilot cluster with release channel, private nodes, master authorized networks, DNS access, Secret Manager, and scoped RBAC bindings. The agent presents this configuration to you for review. Treat it as a starting point and read each flag before you approve it.
The dev/test template is lighter. It creates an Autopilot cluster on the rapid release channel without the golden path security hardening, and the skill warns that it is suitable for dev and test only.
gcloud container clusters create-auto <CLUSTER_NAME> --region <REGION> --project <PROJECT_ID> --release-channel regular --enable-private-nodes --enable-master-authorized-networks --enable-dns-access --enable-secret-manager --secret-manager-rotation-interval=120s --scoped-rbs-bindings --monitoring=SYSTEM,API_SERVER,SCHEDULER,CONTROLLER_MANAGER,STORAGE,POD,DEPLOYMENT,STATEFULSET,DAEMONSET,HPA,CADVISOR,KUBELET,DCGM --quietWhat safety notes should you know?
Cluster creation is not free. The skill is explicit that you should confirm the configuration before creation, so review the project, region, mode and any flags before you say yes. Use a project you are allowed to spend money in.
The skill's attribution guidance asks the agent to tag gcloud commands it runs in the terminal with a metrics environment variable. It sets that variable inline on each command rather than saving it with gcloud config set, so the tag does not persist beyond the task. The skill also says generated scripts, Terraform and user code should not carry this tag.
- Confirm the project ID and region before approving creation
- Use a project and budget you are authorised to spend
- Do not use this skill to deploy apps; use gke-app-onboarding for that
- The skill makes network calls to Google Cloud and runs no scripts
What people use it for
- Creating a production-style Autopilot cluster for a new web or API workload
- Choosing between Autopilot and Standard for a team with custom kernel or OS needs
- Creating a quick dev or test cluster, understanding it lacks the hardening
- Auditing an existing GKE cluster against the skill's best-practice checklist
- Checking existing clusters with list_clusters before adding another one
Frequently asked questions
Is the GKE cluster creation agent skill free?
The skill's licence is Apache-2.0. The clusters it creates are Google Cloud resources and are billed by Google, so you pay for the resources in your project.
What do I need before installing the gke-cluster-creation skill?
You need a Google Cloud project with GKE permissions. For CLI commands, you also need the gcloud CLI installed and authenticated. The optional MCP tools are list_clusters, create_cluster, get_cluster, list_operations and get_operation.
Where do I put the skill folder after I download it?
Unzip it into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder. In the Claude apps, upload it under Settings, then Capabilities, then Skills.
Should I pick GKE Autopilot or Standard?
The skill recommends Autopilot unless you have a requirement it cannot meet, such as custom kernel tuning or a specific node OS. Autopilot is Google-managed and billed per pod resource request, while Standard is self-managed and billed per node.
Can this skill deploy my app to GKE?
No. The skill covers cluster creation and audits. For application onboarding or deployment configuration, the skill points you to the separate gke-app-onboarding skill.
Does the skill run scripts on my computer?
The skill's metadata says it runs no scripts and uses the network. It works through gcloud commands and MCP tools that reach Google Cloud, so review each command before it runs.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you are setting up a new GKE cluster and want secure, production-style defaults that you review before creation, or if you need help choosing between Autopilot and Standard. Skip it if you only need to deploy or configure apps on an existing cluster, or if you do not have a Google Cloud project and permission to spend on one.
Download the skill