Skill UI

This bundle is designed to collect comprehensive diagnostics for CoreWeave clusters, including GPU node status, Kubernetes Pod logs, event history, and API connectivity. It aggregates critical cluster-level resource allocation and failure data into a single, actionable archive. Use it when preparing a support ticket to expedite the diagnosis of complex infrastructure issues like stuck pods, OOM errors, or resource bottlenecks.

This skill provides comprehensive guidance for deploying and managing GPU-accelerated AI inference services on CoreWeave Kubernetes (CKS). It covers best practices including containerization using NVIDIA CUDA base images, configuring specific GPU resource limits (A100/H100), setting up robust health checks, and executing controlled rolling updates. Ideal for managing multi-model inference and scaling demanding AI workloads in a cloud environment.

This skill outlines how to implement and manage Role-Based Access Control (RBAC) and namespace isolation for multi-team GPU workloads running on CoreWeave's Kubernetes platform. It provides guidance on setting strict resource quotas, preventing noisy-neighbor issues, and defining granular roles (e.g., ML Engineer, Inference Operator). This is essential for maintaining resource isolation, ensuring compliance (SOC 2, HIPAA), and managing access across large-scale cloud deployments.

This guide demonstrates how to deploy GPU workloads on the CoreWeave platform using Kubernetes (kubectl). It covers deploying vLLM inference services for large language model (LLM) testing and running batch CUDA jobs for benchmarking. This is ideal for users running their first GPU workloads, testing inference pipelines, or verifying cluster access on bare-metal GPU nodes (A100, H100).

This runbook provides structured steps for responding to critical production incidents on the CoreWeave platform. Use it when dealing with GPU workload failures, inference service outages, or general Kubernetes resource issues. It guides users through checking pod status, node health, and diagnosing common model loading errors to ensure rapid service recovery.

This guide provides step-by-step instructions to configure access to the CoreWeave Kubernetes Service (CKS). It covers downloading the kubeconfig file, setting API tokens, and verifying GPU node access using standard `kubectl` commands. This is essential for deploying AI/ML workloads on bare-metal GPU infrastructure.

This guide provides a deep dive into migrating machine learning workloads, including inference services and training pipelines, from major hyperscalers (AWS, GCP, Azure) to the CoreWeave GPU cloud. It covers cost comparisons, containerization, Kubernetes YAML adaptation, and a phased deployment strategy, helping users ensure a smooth, cost-effective, and optimized transition.

This guide details setting up and managing CoreWeave GPU cloud infrastructure across Development, Staging, and Production environments. It ensures strict resource separation using dedicated Kubernetes namespaces for GPU quotas and scaling policies. Use this when deploying complex ML/AI workloads requiring robust environment isolation and controlled promotion workflows.

This skill provides comprehensive observability for CoreWeave's GPU-intensive workloads running on Kubernetes. It monitors critical operational metrics, including GPU utilization, memory usage, inference latency (p99), and job completion rates. Use it for setting up DCGM metrics dashboards, configuring proactive alerts for underutilization or Out-Of-Memory (OOM) conditions, and ensuring high reliability for AI/ML deployments.

This guide provides expert strategies for optimizing GPU inference performance on CoreWeave infrastructure. It covers GPU selection based on workload (LLM, image generation, training), advanced techniques like continuous batching (using vLLM), autoscaling setup (HPA), and benchmarking data. Use this to maximize GPU utilization, minimize latency, and optimize large-scale AI model deployment.

A comprehensive checklist for ensuring GPU workloads, such as inference services or model training pipelines, are fully prepared for production deployment on CoreWeave. It covers critical MLOps and DevOps aspects including autoscaling, resource limits, security policies, persistent storage validation, monitoring setup, and rollback procedures.

This skill provides methods for managing GPU quota limits and rate restrictions for CoreWeave cloud services. It includes using `kubectl` to check resource quotas and demonstrates implementing robust request queuing mechanisms using Python's `asyncio` library, ensuring stable and efficient resource allocation for high-concurrency inference endpoints.