
Guides an agent through lowering GKE cluster and workload costs with rightsizing and Spot
- Open source
- 24.4K GitHub stars
- Cloud and infrastructure

Guides an agent through lowering GKE cluster and workload costs with rightsizing and Spot

Finds the right GKE golden base image for building custom node images

Applies GKE production defaults, readiness checklists and config audits for clusters

Deploys and tunes LLM and AI inference servers on GKE with GPUs and TPUs

Generates hardened, GKE-ready Kubernetes YAML manifests for Autopilot and Standard

Plans and sets up namespace isolation, quotas, RBAC and cost tracking for GKE teams

Guides an agent through GKE cluster networking: private clusters, VPC-native setup

Guides an agent through diagnosing GKE nodes stuck in NotReady status

Guides an agent through GKE monitoring, logging, and Managed Prometheus setup

Guides an agent through hardening GKE cluster-level security settings and IAM roles

Orchestrates GKE production readiness reviews across scaling, security, reliability

Guides agents to configure GKE workload reliability with PDBs, health probes and zone

Guides GKE edge networking: Gateway API, Ingress, Cloud Armor, NEGs, PSC and LB

Finds GKE StatefulSets with PDBs but no readiness probes and proposes fixes

Configures GKE persistent disks, Filestore shares and StorageClasses for pods

Diagnoses GKE persistent storage failures like attach, mount, pressure and expansion

Finds GKE StatefulSets missing a matching PodDisruptionBudget and proposes fixes

Plans and validates GKE cluster upgrades, node pool strategies and maintenance runbooks

Diagnoses and fixes GKE Workload Identity errors that stop Pods reaching Google Cloud APIs

Guides an agent through scaling GKE workloads with manual scaling, HPA and VPA

Diagnoses GKE HPA failures like unknown metrics, FailedGet events and stalled scaling

Audits and hardens GKE workload security: network policies, pod standards, gVisor

Diagnoses failing GKE workloads from pod status, events, logs and network checks

Lets an agent submit, monitor and cancel TAO GPU training jobs on Kubernetes clusters

Checks and installs the NVIDIA driver, CUDA and container toolkit for TAO GPU hosts

Audits a Linux host and Kubernetes cluster for Cosmos Data Factory controller readiness

Deploys and verifies an Airflow controller for CDF orchestration on Kubernetes

Guides deploying and configuring NVIDIA DOCA Telemetry Service on BlueField, host or DPF

Checks that a Dynamo deployment's NIXL/UCX/NCCL interconnect is ready for disaggregated

Selects, validates, patches and deploys existing NVIDIA Dynamo Kubernetes recipes