[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6993SPN6JDS7NJD2G8NJAXE":3,"skill-same:01M6993SPN6JDS7NJD2G8NJAXE":77},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":40,"linkOnly":41,"markdownContent":42,"markdownBody":43,"frontMatter":44,"refName":48,"commitSha":49,"skillDirectory":50,"forksCount":51,"repositoryHtmlUrl":52,"repositoryHomepageUrl":6,"ownerHtmlUrl":53,"ownerBlogUrl":54,"ownerDescription":55,"updatedAt":56,"install":57,"resources":70},"01M6993SPN6JDS7NJD2G8NJAXE","gke-cluster-creation",null,"Plans and executes GKE cluster creation, provisioning, and production readiness audits using pre-defined templates (Autopilot, Standard Regional, GPU\u002FAI Inference, AI Hypercompute). Use when creating GKE clusters, provisioning GKE environments, selecting cluster modes, or auditing GKE clusters. Don't use for application onboarding or deployment configuration (use gke-app-onboarding instead).","google\u002Fskills","google","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F1342004?v=4","01M698VYEFDM0VJBZDYYJ2T83S","Google","01M698WR0H76ENHE587T9EJT89",23908,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-cluster-creation\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39],"gke cluster creation","gke","cluster","creation","Clusters","containers","when","standard","modes","for","environments","readiness","app","plans","production readiness","use","provisioning","templates","gpu","deployment","inference","2026-11-10 10:12:30",false,"---\nname: gke-cluster-creation\ndescription: >-\n  Plans and executes GKE cluster creation, provisioning, and production\n  readiness audits using pre-defined templates (Autopilot, Standard Regional,\n  GPU\u002FAI Inference, AI Hypercompute). Use when creating GKE clusters,\n  provisioning GKE environments, selecting cluster modes, or auditing GKE\n  clusters. Don't use for application onboarding or deployment configuration\n  (use gke-app-onboarding instead).\nmetadata:\n  version: \"1.0.1\"\n  category: Containers\n---\n\n# GKE Cluster Creation\n\nThis reference guides creating Google Kubernetes Engine (GKE) clusters by\nproviding a set of best-practice templates and guiding through mode selection\nand customization. The **golden path Autopilot** configuration is the default\nfor all new clusters.\n\n> **MCP Tools:** `list_clusters`, `create_cluster`, `get_cluster`,\n> `list_operations`, `get_operation`\n\n## Attribution\n\n**MUST READ**: Before executing any `gcloud`, REST API, or MCP command, you\nMUST read [Usage Attribution](references\u002Fattribution.md). For commands executed\ndirectly in the terminal (never in generated scripts, Terraform, or user code),\nprefix `gcloud` with\n`CLOUDSDK_METRICS_REQUEST_ATTRIBUTION=\"agent-skills\u002F1.0 (skill:gke-cluster-creation)\"`\nand append `agent-skills\u002F1.0 (skill:gke-cluster-creation)` to `User-Agent` headers.\n\n## Workflow\n\n1.  **Discover context**: Use `list_clusters` to see existing clusters. Use\n    `gcloud config get-value project` if project unknown.\n2.  **Gather inputs**: `project_id`, `location` (region or zone),\n    `cluster_name`, environment type. If missing essential details, ask the user\n    before taking action.\n3.  **Select mode & explain trade-offs**: If the user hasn't specified a\n    template or mode, present the available templates (e.g., Autopilot, Standard\n    Regional, GPU Inference, AI Hypercompute) and explain key trade-offs (Cost\n    vs. Availability, Autopilot vs. Standard node management).\n4.  **Configure networking**: auto-create subnet (default) or bring-your-own.\n5.  **Review golden path settings**: present the default configuration block\n    (`gcloud` command or `create_cluster` JSON payload) and confirm with the\n    user before creation.\n6.  **Create**: Use MCP `create_cluster` tool or `gcloud` CLI.\n7.  **Track**: Use `get_operation` to monitor creation progress.\n8.  **Verify**: Use `get_cluster` with `readMask=\"*\"` to confirm golden path\n    settings applied.\n\n## Mode Selection\n\n| Criteria           | Autopilot (Golden Path)   | Standard                  |\n| ------------------ | ------------------------- | ------------------------- |\n| Node management    | Google-managed            | Self-managed              |\n| Pricing            | Pay per pod resource      | Pay per node (VM)         |\n:                    : request                   :                           :\n| Node customization | Via ComputeClasses        | Full control              |\n| DaemonSets         | Allowed (with             | Full control              |\n:                    : restrictions)             :                           :\n| GPU\u002FTPU            | Supported via             | Supported via node pools  |\n:                    : ComputeClasses            :                           :\n| Best for           | Most production workloads | Kernel tuning, custom OS, |\n:                    :                           : privileged workloads      :\n\n> **Rule**: Default to Autopilot unless the customer has a specific requirement\n> that Autopilot cannot satisfy.\n\n## Best Practices\n\nWhen guiding the user or generating configurations, adhere to these GKE best\npractices:\n\n### Security & Networking\n\n1.  **Private Clusters**: Default to private clusters (`enablePrivateNodes:\n    true`) with a private control plane and restricted public endpoints\n    (`enable-master-authorized-networks`) to minimize attack surface.\n2.  **VPC-Native Networking**: Use VPC-native clusters (`useIpAliases: true` \u002F\n    `--enable-ip-alias`) to enable alias IP ranges and pod-level firewall rules.\n3.  **Workload Identity**: Prefer Workload Identity (`workloadPool:\n    \u003CPROJECT_ID>.svc.id.goog`) for securely granting GKE workloads access to\n    Google Cloud services instead of static service account keys.\n4.  **Shielded GKE Nodes**: Enable Shielded GKE Nodes\n    (`--enable-shielded-nodes`, `--enable-secure-boot`) against rootkits and\n    bootkits.\n5.  **Least Privilege (RBAC)**: Institute strict Role-Based Access Control\n    limits (`scoped-rbs-bindings`).\n\n### Cost Optimization\n\n1.  **Autoscaling**: Enable Cluster Autoscaler and Horizontal\u002FVertical Pod\n    Autoscaler (`--enable-autoscaling`, `--enable-vertical-pod-autoscaling`) to\n    adjust resources based on demand.\n2.  **Right-Sizing & Spot VMs**: Choose appropriate machine types and node\n    counts. Consider Spot VMs (`--spot`) for fault-tolerant, non-critical batch\n    or inference workloads.\n\n### High Availability & Reliability\n\n1.  **Regional Clusters**: Use Regional Clusters for production environments to\n    ensure control plane replication across multiple zones (`--region` instead\n    of `--zone`). *Note: Standard regional creates nodes across 3 zones by\n    default.*\n2.  **Pod Disruption Budgets**: Recommend setting Pod Disruption Budgets for\n    application stability during node maintenance.\n3.  **Release Channels**: Subscribe to a release channel (`REGULAR` or `STABLE`)\n    for automated, safer cluster upgrades.\n\n## Templates\n\n### 1. Golden Path Autopilot (Production)\n\nThis is the default. All settings match\n`..\u002Fgke-golden-path\u002Fassets\u002Fgolden-path-autopilot.yaml`.\n\n**Via gcloud:**\n\n```bash\ngcloud container clusters create-auto \u003CCLUSTER_NAME> \\\n  --region \u003CREGION> \\\n  --project \u003CPROJECT_ID> \\\n  --release-channel regular \\\n  --enable-private-nodes \\\n  --enable-master-authorized-networks \\\n  --enable-dns-access \\\n  --enable-secret-manager \\\n  --secret-manager-rotation-interval=120s \\\n  --scoped-rbs-bindings \\\n  --monitoring=SYSTEM,API_SERVER,SCHEDULER,CONTROLLER_MANAGER,STORAGE,POD,DEPLOYMENT,STATEFULSET,DAEMONSET,HPA,CADVISOR,KUBELET,DCGM \\\n  --quiet\n```\n\n**Via MCP (`create_cluster`):**\n\n```json\n{\n  \"parent\": \"projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\",\n  \"cluster\": {\n    \"name\": \"\u003CCLUSTER_NAME>\",\n    \"autopilot\": { \"enabled\": true },\n    \"privateClusterConfig\": { \"enablePrivateNodes\": true },\n    \"masterAuthorizedNetworksConfig\": {\n      \"privateEndpointEnforcementEnabled\": true\n    },\n    \"releaseChannel\": { \"channel\": \"REGULAR\" },\n    \"secretManagerConfig\": {\n      \"enabled\": true,\n      \"rotationConfig\": { \"enabled\": true, \"rotationInterval\": \"120s\" }\n    },\n    \"rbacBindingConfig\": {\n      \"enableInsecureBindingSystemAuthenticated\": false,\n      \"enableInsecureBindingSystemUnauthenticated\": false\n    }\n  }\n}\n```\n\n### 2. Autopilot Dev\u002FTest\n\nRelaxes some golden path defaults for cost savings and easier access in\nnon-production.\n\n**Via gcloud:**\n\n```bash\ngcloud container clusters create-auto \u003CCLUSTER_NAME> \\\n  --region \u003CREGION> \\\n  --project \u003CPROJECT_ID> \\\n  --release-channel rapid \\\n  --quiet\n```\n\n**Via MCP (`create_cluster`):**\n\n```json\n{\n  \"parent\": \"projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\",\n  \"cluster\": {\n    \"name\": \"\u003CCLUSTER_NAME>\",\n    \"autopilot\": { \"enabled\": true },\n    \"releaseChannel\": { \"channel\": \"RAPID\" }\n  }\n}\n```\n\n> **Warning**: This does not apply golden path security hardening. Suitable for\n> dev\u002Ftest only.\n\n### 3. Standard Regional (High Availability \u002F Custom Requirements)\n\nBest when Autopilot cannot be used (e.g., custom kernel tuning, specific node OS\nrequirements). Creates 3 nodes across zones by default.\n\n**Via gcloud:**\n\n```bash\ngcloud container clusters create \u003CCLUSTER_NAME> \\\n  --region \u003CREGION> \\\n  --project \u003CPROJECT_ID> \\\n  --num-nodes 3 \\\n  --machine-type e2-standard-4 \\\n  --disk-type pd-balanced \\\n  --enable-autoscaling --min-nodes 1 --max-nodes 10 \\\n  --enable-shielded-nodes --enable-secure-boot \\\n  --workload-pool=\u003CPROJECT_ID>.svc.id.goog \\\n  --enable-private-nodes \\\n  --enable-master-authorized-networks \\\n  --enable-vertical-pod-autoscaling \\\n  --enable-dataplane-v2 \\\n  --release-channel regular \\\n  --quiet\n```\n\n**Via MCP (`create_cluster`):**\n\n```json\n{\n  \"parent\": \"projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\",\n  \"cluster\": {\n    \"name\": \"\u003CCLUSTER_NAME>\",\n    \"initialNodeCount\": 3,\n    \"nodeConfig\": {\n      \"machineType\": \"e2-standard-4\",\n      \"diskType\": \"pd-balanced\",\n      \"diskSizeGb\": 100,\n      \"oauthScopes\": [\"https:\u002F\u002Fwww.googleapis.com\u002Fauth\u002Fcloud-platform\"],\n      \"shieldedInstanceConfig\": {\n        \"enableSecureBoot\": true,\n        \"enableIntegrityMonitoring\": true\n      },\n      \"workloadMetadataConfig\": {\n        \"mode\": \"GKE_METADATA\"\n      }\n    },\n    \"privateClusterConfig\": { \"enablePrivateNodes\": true },\n    \"releaseChannel\": { \"channel\": \"REGULAR\" },\n    \"workloadIdentityConfig\": {\n      \"workloadPool\": \"\u003CPROJECT_ID>.svc.id.goog\"\n    }\n  }\n}\n```\n\n### 4. GPU Inference & AI Workloads (L4 \u002F ComputeClass)\n\nBest for: AI\u002FML Inference, small model serving. Can be provisioned via\nAutopilot + ComputeClass or via Standard node pool with `g2-standard-4`\n(`nvidia-l4`). *Note: Requires `g2-standard-4` quota.*\n\n**Autopilot ComputeClass \u002F GIQ approach:**\n\n```bash\n# 1. Create golden path cluster (same as template 1)\ngcloud container clusters create-auto \u003CCLUSTER_NAME> \\\n  --region \u003CREGION> --project \u003CPROJECT_ID> \\\n  --enable-private-nodes --enable-master-authorized-networks \\\n  --enable-dns-access --enable-secret-manager --scoped-rbs-bindings \\\n  --quiet\n\n# 2. Apply GPU ComputeClass (see gke-compute-classes.md)\nkubectl apply -f gpu-compute-class.yaml\n\n# 3. Or use GIQ for inference (see gke-inference.md)\ngcloud container ai profiles manifests create \\\n  --model=gemma-2-9b-it --model-server=vllm --accelerator-type=nvidia-l4 --quiet > inference.yaml\nkubectl apply -f inference.yaml\n```\n\n**Standard Node Pool approach via MCP (`create_cluster`):**\n\n```json\n{\n  \"parent\": \"projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\",\n  \"cluster\": {\n    \"name\": \"\u003CCLUSTER_NAME>\",\n    \"initialNodeCount\": 1,\n    \"nodeConfig\": {\n      \"machineType\": \"g2-standard-4\",\n      \"accelerators\": [\n        {\n          \"acceleratorCount\": \"1\",\n          \"acceleratorType\": \"nvidia-l4\"\n        }\n      ],\n      \"diskSizeGb\": 100,\n      \"oauthScopes\": [\"https:\u002F\u002Fwww.googleapis.com\u002Fauth\u002Fcloud-platform\"]\n    }\n  }\n}\n```\n\n### 5. AI Hypercompute (A3 HighGPU \u002F Large Model Serving)\n\nBest for: Large-scale LLM \u002F AI model training and hypercompute inference. *Note:\nHigh hourly cost and strict quota requirements (`a3-highgpu-8g` \u002F\n`nvidia-h100-80gb-hbm3`).*\n\n**Via gcloud:**\n\n```bash\ngcloud container clusters create \u003CCLUSTER_NAME> \\\n  --region \u003CREGION> \\\n  --project \u003CPROJECT_ID> \\\n  --num-nodes 1 \\\n  --machine-type a3-highgpu-8g \\\n  --accelerator type=nvidia-h100-80gb-hbm3,count=8 \\\n  --disk-size 200 \\\n  --scopes https:\u002F\u002Fwww.googleapis.com\u002Fauth\u002Fcloud-platform \\\n  --workload-pool=\u003CPROJECT_ID>.svc.id.goog \\\n  --release-channel regular \\\n  --quiet\n```\n\n**Via MCP (`create_cluster`):**\n\n```json\n{\n  \"parent\": \"projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\",\n  \"cluster\": {\n    \"name\": \"\u003CCLUSTER_NAME>\",\n    \"initialNodeCount\": 1,\n    \"nodeConfig\": {\n      \"machineType\": \"a3-highgpu-8g\",\n      \"accelerators\": [\n        {\n          \"acceleratorCount\": \"8\",\n          \"acceleratorType\": \"nvidia-h100-80gb-hbm3\"\n        }\n      ],\n      \"diskSizeGb\": 200,\n      \"oauthScopes\": [\"https:\u002F\u002Fwww.googleapis.com\u002Fauth\u002Fcloud-platform\"]\n    }\n  }\n}\n```\n\n## Instructions\n\n-   **ALWAYS** ask for `project_id` if not in context.\n-   **ALWAYS** ask for `region` (or location).\n-   **ALWAYS** ask for a unique `cluster_name`.\n-   **DEFAULT** to golden path Autopilot unless customer specifies otherwise or\n    has custom node\u002Fkernel\u002Fhypercompute requirements.\n-   **ALWAYS WARN** when deviating to GKE Standard, highlighting that it\n    deviates from the golden path and explaining the added\n    operational\u002Fmanagement overhead (manually managing node pools, upgrades, and\n    autoscaling).\n-   **EXPLAIN TRADE-OFFS** when presenting templates or mode choices to the user\n    if they haven't specified one (e.g., Autopilot vs Standard, Cost vs\n    Availability).\n-   **PRESENT THE CONFIGURATION** block (`gcloud` command or JSON payload) and\n    ask for confirmation before calling any creation tool.\n-   **WARN** about Day-0 decisions (networking, private nodes) that are hard to\n    change later.\n-   **WARN** explicitly about cost and quota requirements when the user selects\n    GPU (`g2-standard-4`, `a3-highgpu-8g`), TPU, or multi-region\u002Fregional\n    clusters (`--region` defaults to 3 zones).\n-   When using MCP `create_cluster`, the `cluster.name` parameter should be the\n    **short name** (e.g., `my-cluster`), not the full resource path\n    (`projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\u002Fclusters\u002F\u003CCLUSTER_NAME>`). The\n    `parent` parameter defines the scope\n    (`projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>`).\n","# GKE Cluster Creation\n\nThis reference guides creating Google Kubernetes Engine (GKE) clusters by\nproviding a set of best-practice templates and guiding through mode selection\nand customization. The **golden path Autopilot** configuration is the default\nfor all new clusters.\n\n> **MCP Tools:** `list_clusters`, `create_cluster`, `get_cluster`,\n> `list_operations`, `get_operation`\n\n## Attribution\n\n**MUST READ**: Before executing any `gcloud`, REST API, or MCP command, you\nMUST read [Usage Attribution](references\u002Fattribution.md). For commands executed\ndirectly in the terminal (never in generated scripts, Terraform, or user code),\nprefix `gcloud` with\n`CLOUDSDK_METRICS_REQUEST_ATTRIBUTION=\"agent-skills\u002F1.0 (skill:gke-cluster-creation)\"`\nand append `agent-skills\u002F1.0 (skill:gke-cluster-creation)` to `User-Agent` headers.\n\n## Workflow\n\n1.  **Discover context**: Use `list_clusters` to see existing clusters. Use\n    `gcloud config get-value project` if project unknown.\n2.  **Gather inputs**: `project_id`, `location` (region or zone),\n    `cluster_name`, environment type. If missing essential details, ask the user\n    before taking action.\n3.  **Select mode & explain trade-offs**: If the user hasn't specified a\n    template or mode, present the available templates (e.g., Autopilot, Standard\n    Regional, GPU Inference, AI Hypercompute) and explain key trade-offs (Cost\n    vs. Availability, Autopilot vs. Standard node management).\n4.  **Configure networking**: auto-create subnet (default) or bring-your-own.\n5.  **Review golden path settings**: present the default configuration block\n    (`gcloud` command or `create_cluster` JSON payload) and confirm with the\n    user before creation.\n6.  **Create**: Use MCP `create_cluster` tool or `gcloud` CLI.\n7.  **Track**: Use `get_operation` to monitor creation progress.\n8.  **Verify**: Use `get_cluster` with `readMask=\"*\"` to confirm golden path\n    settings applied.\n\n## Mode Selection\n\n| Criteria           | Autopilot (Golden Path)   | Standard                  |\n| ------------------ | ------------------------- | ------------------------- |\n| Node management    | Google-managed            | Self-managed              |\n| Pricing            | Pay per pod resource      | Pay per node (VM)         |\n:                    : request                   :                           :\n| Node customization | Via ComputeClasses        | Full control              |\n| DaemonSets         | Allowed (with             | Full control              |\n:                    : restrictions)             :                           :\n| GPU\u002FTPU            | Supported via             | Supported via node pools  |\n:                    : ComputeClasses            :                           :\n| Best for           | Most production workloads | Kernel tuning, custom OS, |\n:                    :                           : privileged workloads      :\n\n> **Rule**: Default to Autopilot unless the customer has a specific requirement\n> that Autopilot cannot satisfy.\n\n## Best Practices\n\nWhen guiding the user or generating configurations, adhere to these GKE best\npractices:\n\n### Security & Networking\n\n1.  **Private Clusters**: Default to private clusters (`enablePrivateNodes:\n    true`) with a private control plane and restricted public endpoints\n    (`enable-master-authorized-networks`) to minimize attack surface.\n2.  **VPC-Native Networking**: Use VPC-native clusters (`useIpAliases: true` \u002F\n    `--enable-ip-alias`) to enable alias IP ranges and pod-level firewall rules.\n3.  **Workload Identity**: Prefer Workload Identity (`workloadPool:\n    \u003CPROJECT_ID>.svc.id.goog`) for securely granting GKE workloads access to\n    Google Cloud services instead of static service account keys.\n4.  **Shielded GKE Nodes**: Enable Shielded GKE Nodes\n    (`--enable-shielded-nodes`, `--enable-secure-boot`) against rootkits and\n    bootkits.\n5.  **Least Privilege (RBAC)**: Institute strict Role-Based Access Control\n    limits (`scoped-rbs-bindings`).\n\n### Cost Optimization\n\n1.  **Autoscaling**: Enable Cluster Autoscaler and Horizontal\u002FVertical Pod\n    Autoscaler (`--enable-autoscaling`, `--enable-vertical-pod-autoscaling`) to\n    adjust resources based on demand.\n2.  **Right-Sizing & Spot VMs**: Choose appropriate machine types and node\n    counts. Consider Spot VMs (`--spot`) for fault-tolerant, non-critical batch\n    or inference workloads.\n\n### High Availability & Reliability\n\n1.  **Regional Clusters**: Use Regional Clusters for production environments to\n    ensure control plane replication across multiple zones (`--region` instead\n    of `--zone`). *Note: Standard regional creates nodes across 3 zones by\n    default.*\n2.  **Pod Disruption Budgets**: Recommend setting Pod Disruption Budgets for\n    application stability during node maintenance.\n3.  **Release Channels**: Subscribe to a release channel (`REGULAR` or `STABLE`)\n    for automated, safer cluster upgrades.\n\n## Templates\n\n### 1. Golden Path Autopilot (Production)\n\nThis is the default. All settings match\n`..\u002Fgke-golden-path\u002Fassets\u002Fgolden-path-autopilot.yaml`.\n\n**Via gcloud:**\n\n```bash\ngcloud container clusters create-auto \u003CCLUSTER_NAME> \\\n  --region \u003CREGION> \\\n  --project \u003CPROJECT_ID> \\\n  --release-channel regular \\\n  --enable-private-nodes \\\n  --enable-master-authorized-networks \\\n  --enable-dns-access \\\n  --enable-secret-manager \\\n  --secret-manager-rotation-interval=120s \\\n  --scoped-rbs-bindings \\\n  --monitoring=SYSTEM,API_SERVER,SCHEDULER,CONTROLLER_MANAGER,STORAGE,POD,DEPLOYMENT,STATEFULSET,DAEMONSET,HPA,CADVISOR,KUBELET,DCGM \\\n  --quiet\n```\n\n**Via MCP (`create_cluster`):**\n\n```json\n{\n  \"parent\": \"projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\",\n  \"cluster\": {\n    \"name\": \"\u003CCLUSTER_NAME>\",\n    \"autopilot\": { \"enabled\": true },\n    \"privateClusterConfig\": { \"enablePrivateNodes\": true },\n    \"masterAuthorizedNetworksConfig\": {\n      \"privateEndpointEnforcementEnabled\": true\n    },\n    \"releaseChannel\": { \"channel\": \"REGULAR\" },\n    \"secretManagerConfig\": {\n      \"enabled\": true,\n      \"rotationConfig\": { \"enabled\": true, \"rotationInterval\": \"120s\" }\n    },\n    \"rbacBindingConfig\": {\n      \"enableInsecureBindingSystemAuthenticated\": false,\n      \"enableInsecureBindingSystemUnauthenticated\": false\n    }\n  }\n}\n```\n\n### 2. Autopilot Dev\u002FTest\n\nRelaxes some golden path defaults for cost savings and easier access in\nnon-production.\n\n**Via gcloud:**\n\n```bash\ngcloud container clusters create-auto \u003CCLUSTER_NAME> \\\n  --region \u003CREGION> \\\n  --project \u003CPROJECT_ID> \\\n  --release-channel rapid \\\n  --quiet\n```\n\n**Via MCP (`create_cluster`):**\n\n```json\n{\n  \"parent\": \"projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\",\n  \"cluster\": {\n    \"name\": \"\u003CCLUSTER_NAME>\",\n    \"autopilot\": { \"enabled\": true },\n    \"releaseChannel\": { \"channel\": \"RAPID\" }\n  }\n}\n```\n\n> **Warning**: This does not apply golden path security hardening. Suitable for\n> dev\u002Ftest only.\n\n### 3. Standard Regional (High Availability \u002F Custom Requirements)\n\nBest when Autopilot cannot be used (e.g., custom kernel tuning, specific node OS\nrequirements). Creates 3 nodes across zones by default.\n\n**Via gcloud:**\n\n```bash\ngcloud container clusters create \u003CCLUSTER_NAME> \\\n  --region \u003CREGION> \\\n  --project \u003CPROJECT_ID> \\\n  --num-nodes 3 \\\n  --machine-type e2-standard-4 \\\n  --disk-type pd-balanced \\\n  --enable-autoscaling --min-nodes 1 --max-nodes 10 \\\n  --enable-shielded-nodes --enable-secure-boot \\\n  --workload-pool=\u003CPROJECT_ID>.svc.id.goog \\\n  --enable-private-nodes \\\n  --enable-master-authorized-networks \\\n  --enable-vertical-pod-autoscaling \\\n  --enable-dataplane-v2 \\\n  --release-channel regular \\\n  --quiet\n```\n\n**Via MCP (`create_cluster`):**\n\n```json\n{\n  \"parent\": \"projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\",\n  \"cluster\": {\n    \"name\": \"\u003CCLUSTER_NAME>\",\n    \"initialNodeCount\": 3,\n    \"nodeConfig\": {\n      \"machineType\": \"e2-standard-4\",\n      \"diskType\": \"pd-balanced\",\n      \"diskSizeGb\": 100,\n      \"oauthScopes\": [\"https:\u002F\u002Fwww.googleapis.com\u002Fauth\u002Fcloud-platform\"],\n      \"shieldedInstanceConfig\": {\n        \"enableSecureBoot\": true,\n        \"enableIntegrityMonitoring\": true\n      },\n      \"workloadMetadataConfig\": {\n        \"mode\": \"GKE_METADATA\"\n      }\n    },\n    \"privateClusterConfig\": { \"enablePrivateNodes\": true },\n    \"releaseChannel\": { \"channel\": \"REGULAR\" },\n    \"workloadIdentityConfig\": {\n      \"workloadPool\": \"\u003CPROJECT_ID>.svc.id.goog\"\n    }\n  }\n}\n```\n\n### 4. GPU Inference & AI Workloads (L4 \u002F ComputeClass)\n\nBest for: AI\u002FML Inference, small model serving. Can be provisioned via\nAutopilot + ComputeClass or via Standard node pool with `g2-standard-4`\n(`nvidia-l4`). *Note: Requires `g2-standard-4` quota.*\n\n**Autopilot ComputeClass \u002F GIQ approach:**\n\n```bash\n# 1. Create golden path cluster (same as template 1)\ngcloud container clusters create-auto \u003CCLUSTER_NAME> \\\n  --region \u003CREGION> --project \u003CPROJECT_ID> \\\n  --enable-private-nodes --enable-master-authorized-networks \\\n  --enable-dns-access --enable-secret-manager --scoped-rbs-bindings \\\n  --quiet\n\n# 2. Apply GPU ComputeClass (see gke-compute-classes.md)\nkubectl apply -f gpu-compute-class.yaml\n\n# 3. Or use GIQ for inference (see gke-inference.md)\ngcloud container ai profiles manifests create \\\n  --model=gemma-2-9b-it --model-server=vllm --accelerator-type=nvidia-l4 --quiet > inference.yaml\nkubectl apply -f inference.yaml\n```\n\n**Standard Node Pool approach via MCP (`create_cluster`):**\n\n```json\n{\n  \"parent\": \"projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\",\n  \"cluster\": {\n    \"name\": \"\u003CCLUSTER_NAME>\",\n    \"initialNodeCount\": 1,\n    \"nodeConfig\": {\n      \"machineType\": \"g2-standard-4\",\n      \"accelerators\": [\n        {\n          \"acceleratorCount\": \"1\",\n          \"acceleratorType\": \"nvidia-l4\"\n        }\n      ],\n      \"diskSizeGb\": 100,\n      \"oauthScopes\": [\"https:\u002F\u002Fwww.googleapis.com\u002Fauth\u002Fcloud-platform\"]\n    }\n  }\n}\n```\n\n### 5. AI Hypercompute (A3 HighGPU \u002F Large Model Serving)\n\nBest for: Large-scale LLM \u002F AI model training and hypercompute inference. *Note:\nHigh hourly cost and strict quota requirements (`a3-highgpu-8g` \u002F\n`nvidia-h100-80gb-hbm3`).*\n\n**Via gcloud:**\n\n```bash\ngcloud container clusters create \u003CCLUSTER_NAME> \\\n  --region \u003CREGION> \\\n  --project \u003CPROJECT_ID> \\\n  --num-nodes 1 \\\n  --machine-type a3-highgpu-8g \\\n  --accelerator type=nvidia-h100-80gb-hbm3,count=8 \\\n  --disk-size 200 \\\n  --scopes https:\u002F\u002Fwww.googleapis.com\u002Fauth\u002Fcloud-platform \\\n  --workload-pool=\u003CPROJECT_ID>.svc.id.goog \\\n  --release-channel regular \\\n  --quiet\n```\n\n**Via MCP (`create_cluster`):**\n\n```json\n{\n  \"parent\": \"projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\",\n  \"cluster\": {\n    \"name\": \"\u003CCLUSTER_NAME>\",\n    \"initialNodeCount\": 1,\n    \"nodeConfig\": {\n      \"machineType\": \"a3-highgpu-8g\",\n      \"accelerators\": [\n        {\n          \"acceleratorCount\": \"8\",\n          \"acceleratorType\": \"nvidia-h100-80gb-hbm3\"\n        }\n      ],\n      \"diskSizeGb\": 200,\n      \"oauthScopes\": [\"https:\u002F\u002Fwww.googleapis.com\u002Fauth\u002Fcloud-platform\"]\n    }\n  }\n}\n```\n\n## Instructions\n\n-   **ALWAYS** ask for `project_id` if not in context.\n-   **ALWAYS** ask for `region` (or location).\n-   **ALWAYS** ask for a unique `cluster_name`.\n-   **DEFAULT** to golden path Autopilot unless customer specifies otherwise or\n    has custom node\u002Fkernel\u002Fhypercompute requirements.\n-   **ALWAYS WARN** when deviating to GKE Standard, highlighting that it\n    deviates from the golden path and explaining the added\n    operational\u002Fmanagement overhead (manually managing node pools, upgrades, and\n    autoscaling).\n-   **EXPLAIN TRADE-OFFS** when presenting templates or mode choices to the user\n    if they haven't specified one (e.g., Autopilot vs Standard, Cost vs\n    Availability).\n-   **PRESENT THE CONFIGURATION** block (`gcloud` command or JSON payload) and\n    ask for confirmation before calling any creation tool.\n-   **WARN** about Day-0 decisions (networking, private nodes) that are hard to\n    change later.\n-   **WARN** explicitly about cost and quota requirements when the user selects\n    GPU (`g2-standard-4`, `a3-highgpu-8g`), TPU, or multi-region\u002Fregional\n    clusters (`--region` defaults to 3 zones).\n-   When using MCP `create_cluster`, the `cluster.name` parameter should be the\n    **short name** (e.g., `my-cluster`), not the full resource path\n    (`projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>\u002Fclusters\u002F\u003CCLUSTER_NAME>`). The\n    `parent` parameter defines the scope\n    (`projects\u002F\u003CPROJECT_ID>\u002Flocations\u002F\u003CREGION>`).\n",{"name":5,"metadata":45,"description":7},{"version":46,"category":47},"1.0.1","Containers","main","bb9b3f464aae37af360d079c1d863524955b0363","skills\u002Fcloud\u002Fgke-cluster-creation",1994,"https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills","https:\u002F\u002Fgithub.com\u002Fgoogle","https:\u002F\u002Fopensource.google\u002F","Google ❤️ Open Source","2026-11-12 00:37:41",{"source":8,"skillName":5,"commands":58,"prompt":68,"zipAvailable":69},[59,62,65],{"tool":60,"command":61},"npx","npx skills add google\u002Fskills --skill gke-cluster-creation",{"tool":63,"command":64},"bunx","bunx skills add google\u002Fskills --skill gke-cluster-creation",{"tool":66,"command":67},"pnpm","pnpm dlx skills add google\u002Fskills --skill gke-cluster-creation","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-cluster-creation\u002FSKILL.md",true,[71],{"path":72,"resourceType":73,"resourceTypeDesc":74,"sizeBytes":75,"sourceHtmlUrl":76},"references\u002Fattribution.md","REFERENCE","参考文档",1900,"https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fgke-cluster-creation\u002Freferences\u002Fattribution.md",[78,116,147,176,203,227],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":92,"publishedAt":40},"01M6998S2EZKD3GEJ7F943RZTE","sign-in-with-google-web","Implement, configure, and secure Sign In With Google (SiwG) using Google Identity Services (GIS \u002F `https:\u002F\u002Faccounts.google.com\u002Fgsi\u002Fclient`) across web architectures. Use when creating Google sign-in buttons, implementing Google One Tap with FedCM, integrating GIS in React\u002FNext.js\u002FAngular\u002FHTML, verifying ID tokens on backend runtimes (Python\u002FNode.js\u002FGo\u002FJava), enforcing Google Workspace domain restrictions (`hd`), securing client-side ID tokens via WebCrypto nonces or encrypted IndexedDB, embedding in cross-origin iframes via the Intermediate Iframe API (`https:\u002F\u002Faccounts.google.com\u002Fgsi\u002Fintermediate`), configuring Content Security Policy (CSP), COOP, and Permissions Policy (`identity-credentials-get`) headers, or handling sign-out and revocation. Don't use for legacy Google Sign-In (`gapi.auth2`), Google Cloud IAM service accounts, or Android\u002FiOS native Credential Manager.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fidentity\u002Fsign-in-with-google-web\u002FSKILL.md",[84,89],{"categoryNo":85,"categoryName":86,"taxonomyType":87,"taxonomyTypeDesc":88},"01JBSEEDCAT000000000000009","前端开发","SCENARIO","使用场景",{"categoryNo":90,"categoryName":91,"taxonomyType":87,"taxonomyTypeDesc":88},"01JBSEEDCAT000000000000010","后端与接口",[9,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115],"sign in with google web","sign","with","web","android","go","security","google workspace","google cloud","implement","policy","python","implementing","react","iam","permissions","buttons","secure","manager","identity","architectures","domain","runtimes",{"skillNo":117,"name":118,"nameZh":6,"description":119,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":120,"categories":121,"keywords":131,"publishedAt":40},"01M6998PHPDCAD684VW5WRG83D","dpop-adoption","Implement and debug OAuth 2.0 DPoP (RFC 9449) refresh token sender-constraining for WebCrypto, Node.js ES6, and browser runtimes integrating with Google's OAuth platform. Use when configuring non-extractable asymmetric key pairs (P-256), generating DPoP Proof JWTs for authorization code exchange and token refresh, or handling 400 use_dpop_nonce challenge retry loops at oauth2.googleapis.com\u002Ftoken. Don't use for unconstrained OAuth 2.0 flows (where refresh tokens are not bound to a client key pair), or for Google Cloud IAM \u002F service account authentication.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fidentity\u002Fdpop-adoption\u002FSKILL.md",[122,123,126],{"categoryNo":90,"categoryName":91,"taxonomyType":87,"taxonomyTypeDesc":88},{"categoryNo":124,"categoryName":125,"taxonomyType":87,"taxonomyTypeDesc":88},"01JBSEEDCAT000000000000023","Skill与智能体开发",{"categoryNo":127,"categoryName":128,"taxonomyType":129,"taxonomyTypeDesc":130},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[132,133,9,101,134,135,136,137,138,139,107,140,141,142,143,144,145,112,115,146,25],"dpop adoption","adoption","es6","retry","authentication","flows","challenge","code","browser","loops","at","node","token","pair","platform",{"skillNo":148,"name":149,"nameZh":6,"description":150,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":151,"categories":152,"keywords":159,"publishedAt":40},"01M6998M72TC73EEAQH4XC7EXW","retrieving-developer-knowledge","Searches, retrieves, and synthesizes official Google developer documentation across Google Cloud, AI\u002FGemini, Android, Chrome, Web, Flutter, Go, Firebase, and other Google developer platforms. Integrates with the Developer Knowledge MCP server (search_documents, get_documents, answer_query) or the Developer Knowledge REST API fallback. Use when searching for gcloud CLI commands, API syntax, IAM permissions, official documentation, architectural comparisons, or product choice overviews. Don't use for local filesystem lookups or non-Google documentation.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fdevelopers\u002Fretrieving-developer-knowledge\u002FSKILL.md",[153,154,157,158],{"categoryNo":90,"categoryName":91,"taxonomyType":87,"taxonomyTypeDesc":88},{"categoryNo":155,"categoryName":156,"taxonomyType":87,"taxonomyTypeDesc":88},"01JBSEEDCAT000000000000021","资料检索与研究",{"categoryNo":124,"categoryName":125,"taxonomyType":87,"taxonomyTypeDesc":88},{"categoryNo":127,"categoryName":128,"taxonomyType":129,"taxonomyTypeDesc":130},[160,161,162,163,101,164,165,166,167,168,169,170,171,25,172,173,174,107,175,108],"retrieving developer knowledge","knowledge","developer","MCP","mcp server","official","developer documentation","get","search","local","choice","documents","query","api","across","documentation",{"skillNo":177,"name":178,"nameZh":6,"description":179,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":180,"categories":181,"keywords":182,"publishedAt":40},"01M6998J2CZEXS9BES9S3WWFKC","migrate-to-rust","Clones, analyzes, plans, and migrates legacy codebases (C++, Fortran, Java, Python, C, Shell) from a GitHub URL or local directory to a modern, high-performance, deterministic, 100% Safe Rust architecture, tracking live telemetry across every task bucket, test failure\u002Ffix cycle, and Profile-Guided Optimization (PGO) stage, and producing a first-person AI storytelling report and interactive HTML infographic. Use when migrating a GitHub repository or local codebase to Rust, porting legacy C++, Fortran, Python, or Java workloads to memory-safe Rust, designing Struct-of-Arrays (SoA) or SIMD Rust architectures, or optimizing Rust ports with LLVM PGO. Don't use for unrelated non-Rust refactoring or single-line syntax questions.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fdevelopers\u002Fmigrate-to-rust\u002FSKILL.md",[],[183,184,185,186,187,188,189,190,104,191,192,193,194,195,196,169,197,198,199,25,200,201,174,202],"migrate to rust","rust","migrate","github","memory","performance","directory","safe","ports","llvm","soa","cycle","porting","fix","arrays","infographic","high-performance","URL","architecture","unrelated",{"skillNo":204,"name":205,"nameZh":6,"description":206,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":207,"categories":208,"keywords":212,"publishedAt":40},"01M6998GCC3K29WZ82KXMB9E02","finding-google-skills","Google platform decision and setup guidance, loaded on demand from Google's skill catalog. Use when a developer is choosing or setting up part of their stack, such as where to run a service, a database, storage, messaging, authentication, analytics, ads, or AI model serving, and a Google product is a reasonable candidate - whether or not a vendor is named - or when a request names a Google product or API. Brings in the matching Google skill so the answer can weigh Google options, their trade-offs, and when they are not the right fit. Skip when the stack is already settled on another provider and no Google product is named, or the task involves no platform choice.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fdevelopers\u002Ffinding-google-skills\u002FSKILL.md",[209,210,211],{"categoryNo":90,"categoryName":91,"taxonomyType":87,"taxonomyTypeDesc":88},{"categoryNo":124,"categoryName":125,"taxonomyType":87,"taxonomyTypeDesc":88},{"categoryNo":127,"categoryName":128,"taxonomyType":129,"taxonomyTypeDesc":130},[9,213,214,215,216,146,217,218,219,220,221,136,222,223,224,170,34,225,25,226,173],"finding google skills","storage","as","messaging","model","trade-offs","provider","answer","fit","catalog","up","decision","database","candidate",{"skillNo":228,"name":229,"nameZh":6,"description":230,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":231,"categories":232,"keywords":235,"publishedAt":40},"01M6998DWNHTH0BS6W1CY8ZXMP","workload-manager-basics","Use this skill to manage Google Cloud Workload Manager evaluations, rules, scanned resources, and validation results by using public client libraries and the REST API. Use when you need to inspect workload best-practice rules, create and run evaluations for Google Cloud general best practices, SAP, SQL Server, or custom organizational rules, review violations, export results to BigQuery, or automate Workload Manager through client libraries because no service-specific public CLI or MCP server is available. Don't use for general Google Compute Engine instance management, VPC configuration, or standard IAM auditing.","https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fcloud\u002Fworkload-manager-basics\u002FSKILL.md",[233,234],{"categoryNo":90,"categoryName":91,"taxonomyType":87,"taxonomyTypeDesc":88},{"categoryNo":124,"categoryName":125,"taxonomyType":87,"taxonomyTypeDesc":88},[236,237,111,238,9,163,239,101,164,240,241,242,243,25,244,173,107,245,246,247,248],"workload manager basics","workload","basics","cli","review","vpc","resources","inspect","you","create","libraries","server","run"]