Skill UI

A comprehensive guide and set of scripts to maximize cost savings in Kubernetes environments using CAST AI. Learn to implement aggressive spot instance strategies, perform workload right-sizing, utilize cluster hibernation for non-production environments, and analyze monthly savings reports to significantly reduce cloud compute expenditure.

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 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 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.

Provides comprehensive guidelines for securing CoreWeave cloud deployments on Kubernetes. Covers critical security areas including credential management, implementing Role-Based Access Control (RBAC), enforcing network isolation policies, and handling sensitive secrets for ML model access. Essential for hardening GPU workloads and protecting data in cloud environments.

This skill provides a comprehensive solution for deploying containerized applications to Oracle Cloud Infrastructure (OCI). It supports two primary deployment paths: OKE (Oracle Kubernetes Engine) for full Kubernetes orchestration, and Container Instances for a simpler, serverless alternative. Use this skill for automating the entire lifecycle, including pushing images to OCIR, provisioning OKE clusters via Python SDK, and executing Kubernetes deployments using kubectl.

This advanced skill automates the entire lifecycle of cloud infrastructure deployment. Instead of manual documentation review, it researches current best practices across multi-cloud platforms (AWS, GCP, Azure), synthesizes a coherent deployment strategy, and generates production-grade Infrastructure as Code (IaC). It outputs comprehensive artifacts, including Terraform modules, Kubernetes manifests, CI/CD pipelines, monitoring setups, and cost estimates, ensuring secure and optimized deployment.

This guide details the process of deploying and configuring the WizTelemetry Data Pipeline using Vector. It is essential for establishing data collection, transformation, and routing capabilities for comprehensive observability data, including logs, auditing records, events, and notifications within a Kubernetes environment.

A specialized skill for building and auditing Kubernetes Operators. It focuses on the core principle of the declarative reconcile loop, helping developers identify and fix common operational bugs such as status drift, missing finalizers, and incorrect RBAC. Includes tools to validate CRD specifications, lint controller code, and audit capability levels against industry best practices.

This skill provides comprehensive operational capabilities for managing FrontendExtension resources in a Kubernetes cluster. It covers the entire lifecycle, including resource creation, package generation, artifact inspection, publishing, unpublishing, deletion, and deep debugging of package and publish controller behaviors. Use it for troubleshooting FE status, managing deployment pipelines, or interacting with the FE API.