Google

Configure GKE ComputeClasses, which define preferred computing resources for workloads. Set fallback options between Spot and on-demand VMs, target GPUs or machine families, control class usage, and investigate pods waiting for suitable nodes.

• Configuring fallback from Spot VMs to on-demand capacity
• Selecting GPUs, TPUs, or machine families
• Restricting ComputeClass editing and workload usage

Reduce Google Kubernetes Engine costs by adjusting CPU and memory requests, setting resource quotas, and choosing Spot VMs or committed use discounts. Match spending decisions to steady usage, variable demand, and workloads that tolerate interruptions.

• Reduce excessive CPU and memory requests
• Set resource quotas for teams sharing a cluster
• Move interruption-tolerant batch jobs to Spot VMs

Deploy AI models on GKE with suitable GPUs or TPUs and model servers. Generate deployment configuration, adjust automatic scaling, and tune inference—the process of producing model responses—for memory use, throughput, and response time.

• Deploying Gemma or Llama on GKE
• Choosing GPUs or TPUs for model serving
• Configuring automatic scaling for LLM requests

Create or update Kubernetes YAML configuration files for GKE Autopilot and Standard. Translates application requirements into manifests covering container permissions, CPU and memory allocation, health checks, storage, external routing, and GPU resources for AI model serving.

• Creating GKE application deployment manifests
• Reviewing container security and resource settings
• Configuring external routes with GKE Gateway API

Plan and configure a shared GKE cluster for multiple teams or environments. Set up namespaces, team permissions, resource limits, and network separation, then use labels and GKE cost allocation to track spending by team.

• Design a GKE cluster shared by multiple teams
• Separate development, staging, and production workloads
• Set team permissions and resource quotas

Diagnose GKE nodes marked NotReady or Unknown using node conditions, events, logs, and resource metrics. The skill gathers evidence, distinguishes temporary transitions from faults, and proposes fixes for a person to apply.

• Investigate persistent NotReady or Unknown node status
• Check why kubelet stopped reporting node status
• Investigate evictions or Pending Pods linked to node health

Plan and configure security across Google Kubernetes Engine clusters. Review access permissions, enable Secret Manager integration, and configure protections such as Shielded Nodes and Binary Authorization to strengthen the platform supporting your applications.

• Review cluster access permissions
• Enable Secret Manager integration
• Configure Shielded Nodes and GKE Sandbox

Review GKE reliability settings for applications running in Kubernetes. Configure availability safeguards during maintenance, health checks, replica counts, graceful shutdown, and distribution across zones and nodes to help workloads handle disruptions.

• Reviewing GKE high availability
• Maintaining Pod availability during upgrades
• Configuring application health checks

Configure how GKE applications receive traffic from the internet or internal networks. The skill provides configuration templates for routing, load balancing, certificates, and access protection, and guides diagnosis of failed health checks, dropped requests, and TLS connection problems.

• Expose a GKE application through Gateway API or Ingress
• Split traffic between application versions during a rollout
• Configure Cloud Armor protection and HTTPS certificates

Plans, executes, and validates Google Kubernetes Engine upgrades for Standard and Autopilot clusters. Creates environment-specific upgrade plans, checklists, and command-based runbooks, with guidance on maintenance timing, node upgrade strategies, workload readiness, and stalled upgrades.

• Planning a GKE version upgrade
• Choosing release channels and maintenance schedules
• Preparing node pool upgrades for databases or GPU workloads

Plan and create Google Kubernetes Engine clusters using templates for production, development, or AI workloads. Compare Google-managed and self-managed node options, review configurations before creation, and check settings for production readiness.

• Create a production Autopilot cluster
• Compare Autopilot and Standard modes
• Prepare a development or test cluster

Diagnose interrupted or repeatedly restarting AI training jobs managed by JobSet on GKE. Correlate restart metrics, node events, host health, and worker logs to distinguish Spot VM preemptions, infrastructure failures, and application crashes, then recommend responses.

• Investigating repeated JobSet restarts
• Checking whether Spot VM preemptions interrupted training
• Tracing recurring failures to a physical host

Inspect GKE TPU Dynamic Slices, diagnose provisioning failures, and validate workload configuration for single slices or JobSet workloads spanning multiple slices. Get cleanup commands, with confirmation required before destructive changes.

• Checking TPU slice lifecycle states
• Diagnosing TPU slice provisioning failures
• Validating single-slice and multi-slice JobSet manifests

Investigate crashes, out-of-memory errors, and device initialization failures on GKE TPU v6e nodes. The skill checks logs and custom metrics collection for conflicting operations, then guides temporary mitigation by disabling an identified collector.

• Investigate vbar_control_agent crashes and memory exhaustion
• Check tpu-device-plugin metrics checksum errors
• Inspect conflicts with custom TPU metrics collectors

Helps configure persistent storage for GKE applications, including disk selection, storage requests, shared files through Filestore, and volume expansion. It also explains how disk compatibility and zone placement affect workload scheduling.

• Creating persistent storage for a GKE application
• Sharing files across multiple Pods with Filestore
• Expanding an existing storage volume

Answers questions about GKE spending using BigQuery billing exports and live cluster metrics. It breaks down costs by project, namespace, or workload and compares requested computing resources with actual usage to investigate cost increases.

• Comparing GKE costs across projects
• Finding the most expensive namespace or workload
• Investigating a sudden increase in cluster costs

Adjust application capacity on GKE through manual scaling, HPA for changing the number of running pods, and VPA for adjusting CPU and memory allocations. Includes configuration templates and a workflow for reviewing resource requests against actual usage.

• Scale application capacity automatically as demand changes
• Review CPU and memory allocations against actual usage
• Change the number of application replicas manually

Finds the base image needed to create a custom GKE node image. It translates requirements such as GKE version, operating system, processor architecture, accelerators, and sandbox settings into a matching image name.

• Finding a base image for custom GKE nodes
• Looking up images for GPU or TPU configurations
• Matching images to COS, Ubuntu, ARM64, or X86_64

Reserve remote Android devices on Google's Developer Device Platform and connect through ADB, Android's device control tool. Check session status and expiration, extend usage time, or cancel sessions and release resources.

• Find and reserve a remote Android device
• Connect to a reserved device through ADB
• Check session status and expiration

Generate a client in your chosen language for live, two-way communication with Gemini Enterprise LiveAPI. Includes token renewal, session recovery, tests, and a demo interface for trying text, audio, video, and transcription.

• Build a client for live conversations with Gemini
• Add token renewal and session recovery
• Test audio and video input through a demo interface

Configure GKE logging, metrics collection, dashboards, and alerts with Cloud Logging, Cloud Monitoring, and managed Prometheus. Diagnose missing measurements, collection failures, permission errors, and alert rule problems to identify appropriate configuration changes.

• Setting up GKE logging and metrics collection
• Investigating missing Prometheus metrics
• Planning dashboards and alerts for cluster health

Assess whether a GKE cluster or application is ready for production. Coordinates specialist skills to review scaling, security, reliability, monitoring, recovery, costs, and maintenance, then reports domain status, a weighted readiness score, and remediation priorities.

• Review a GKE application before its production launch
• Assess production readiness across an entire cluster
• Prioritize fixes using domain status and a readiness score

Diagnose why GKE Pods cannot authenticate to Google Cloud APIs. Inspect cluster settings, service account bindings, resource permissions, and metadata server health, then provide evidence and concrete fixes for a human to apply.

• Investigate Pod authentication failures and 403 errors
• Check Kubernetes service account identity bindings
• Diagnose metadata server connectivity problems

Design Google Cloud infrastructure with Terraform, validate it locally, and assess the planned changes for security, cost, and reliability. The skill then imports templates into Application Design Center, deploys them, investigates failures, and checks the deployed services.

• Designing and locally validating Terraform for GCP
• Assessing infrastructure plans before deployment
• Importing and deploying Application Design Center templates