airunway-aks-setup
🎯SkillThis skill walks you step by step from an Azure Kubernetes Service (AKS) cluster, a group of cloud servers, to a running AI model. It checks your servers and GPUs, installs AI Runway and deploys your first model.
Installation
npx skills add microsoft/azure-skills --skill airunway-aks-setupRun it in a terminal; it asks which AI coding tools to install it for (Claude Code, Codex, Cursor and more).
Uses skills.sh · Node.js 22.20+
After installing
Ask for related work and the agent uses it on its own, or call it yourself.
/airunway-aks-setupInstall by hand
Copy the whole skill folder from GitHub to the place below. Copying only SKILL.md can leave out files the skill uses.
Open the skill folder →- This project
.claude/skills/airunway-aks-setup/ - All projects
~/.claude/skills/airunway-aks-setup/
These paths are for Claude Code. For other tools, the command above puts it in the right place.
What this skill does
Overview
AI Runway is a tool for running AI models on AKS, the Microsoft Azure service that manages groups of servers with Kubernetes. This skill takes you through six steps in order, using kubectl, make and curl. It reports the state of your cluster at each step and asks you before it installs or deploys anything.
Key Features
- Check that you are connected to the right cluster and see which servers and GPUs it has.
- Install the AI Runway controller, the part that manages model deployments on your cluster.
- Find GPU limits early, for example that T4 and V100 cards do not support
bfloat16. - Pick and install an inference provider, the software that runs the model: KAITO, Dynamo or KubeRay.
- Deploy a first model, run a quick test, and resume from any step with
skip-to-step N.
Who is this for?
It is for engineers who already have an AKS cluster and want to serve a language model on GPUs for the first time. If you do not have a cluster yet, the skill sends you to the azure-kubernetes skill first, and it reminds you that an A100-80GB GPU can cost $3 to $5 or more per hour.
Summarized by AI from this skill's SKILL.md. Suggest a better one
SKILL.md
View on GitHub →Set up AI Runway on AKS — from bare cluster to running model. Covers cluster verification, controller install, GPU assessment, provider setup, and first deployment. WHEN: \"setup AI Runway\", \"onboard AKS cluster\", \"install AI Runway\", \"airunway setup\", \"deploy model to AKS\", \"GPU inference on AKS\", \"KAITO setup on AKS\", \"run LLM on AKS\", \"vLLM on AKS\", \"set up model serving on AKS\", \"AI Runway controller\".
AI Runway AKS Setup
This skill walks users from a bare Kubernetes cluster to a running AI model deployment. Follow each step in sequence unless the user provides skip-to-step N to resume from a specific phase.
Cost awareness: GPU node pools incur significant compute charges (A100-80GB can cost $3–5+/hr). Confirm the user understands cost implications before provisioning GPU resources.
Prerequisites
This skill assumes an AKS cluster already exists. If the user does not have a cluster, hand off to the azure-kubernetes skill first to provision one (with a GPU node pool unless CPU-only inference is acceptable), then return here.
Quick Reference
| Property | Value |
|---|---|
| Best for | End-to-end AI Runway onboarding on AKS |
| CLI tools | kubectl, make, curl |
| MCP tools | None |
| Related skills | azure-kubernetes (cluster setup), azure-diagnostics (baseline troubleshooting), optional aks-gpu-inference (existing GPU/inference Day-2 incidents only) |
When to Use This Skill
Use this skill when the user wants to:
- Set up AI Runway on an existing AKS cluster from scratch
- Install the AI Runway controller and CRDs
- Assess GPU hardware compatibility for model deployment
- Choose and install an inference provider (KAITO, Dynamo, KubeRay)
- Deploy their first AI model to AKS via AI Runway
- Resume a partially-complete AI Runway setup from a specific step
GPU setup versus Day-2 incidents
Keep initial GPU node-pool, provider, controller, and model setup in this skill. For an incident affecting an existing GPU or inference deployment, load optional-gpu-day2.md before deciding whether to offer the focused optional add-on.
MCP Tools
This skill uses no MCP tools. All cluster operations are performed directly via kubectl and make.
Rules
- Execute steps in sequence — load the reference for each step as you reach it
- Report cluster state at each step: ✓ healthy, ✗ missing/failed
- Ask for user confirmation before any install or deployment action
- If a step is already complete, report status and skip to the next step
- If the user provides
skip-to-step N, start at step N; assume prior steps are complete
Steps
| # | Step | Reference |
|---|---|---|
| 1 | Cluster Verification — context check, node inventory, GPU detection | step-1-verify.md |
| 2 | Controller Installation — CRD + controller deployment | step-2-controller.md |
| 3 | GPU Assessment — detect GPU models, flag dtype/attention constraints | step-3-gpu.md |
| 4 | Provider Setup — recommend and install inference provider | step-4-provider.md |
| 5 | First Deployment — pick a model, deploy, verify Ready | step-5-deploy.md |
| 6 | Summary — recap, smoke test, next steps | step-6-summary.md |
Error Handling
| Error / Symptom | Likely Cause | Remediation |
|---|---|---|
| No kubeconfig context | Not connected to a cluster | Run az aks get-credentials or equivalent |
| Controller in CrashLoopBackOff | Config or RBAC issue | kubectl logs -n airunway-system -l control-plane=controller-manager --previous |
| Provider not ready | Image pull or RBAC issue | kubectl logs <pod-name> -n <namespace> for the provider pod |
| ModelDeployment stuck in Pending | GPU scheduling failure or provider not ready | kubectl describe modeldeployment <name> -n <namespace> events |
bfloat16 errors at inference | T4 or V100 lacks bfloat16 support | Add --dtype float16 to serving args |
For full error handling and rollback procedures, see troubleshooting.md.
More from this repository10
Helps you build, deploy, test, and manage AI agents and models on Microsoft Foundry, Microsoft's Azure service for AI. It guides the AI through each task with fixed steps.
Helps you find and fix problems in apps running on Azure, such as crashes, slow responses, or connection failures. It checks things in a set order and can work only from logs you provide.
azure-prepare gets a new or existing app ready to deploy to Azure with azd, Microsoft's command-line tool for Azure. It first writes a plan for your approval, then creates the setup files.
azure-deploy puts your app on Azure after the azure-prepare and azure-validate skills have prepared and checked it. It runs the deployment, recovers from errors, and gives you the live web addresses.
azure-validate checks your Azure app before you deploy it, so you can catch setup and permission problems early. It runs after azure-prepare and before azure-deploy.
azure-resource-lookup finds and lists your Azure resources, such as web apps, virtual machines, and storage accounts, across your subscriptions. It only reads data and never changes anything.
azure-resource-visualizer draws a diagram of an Azure resource group that shows how each resource connects to the others. It saves the diagram and notes in a markdown file.
Lets Claude search and summarize large amounts of logs and app data stored in Azure Data Explorer. You ask what you want to know, and it writes and runs the queries for you.
Helps you send all requests to your AI models and tools through Azure API Management, so you can control cost, block harmful content and limit traffic. It includes Azure command-line steps for setup and testing.
Helps you find and fix errors in apps that send and receive messages through Azure Event Hubs or Service Bus. It checks service health and Microsoft docs to suggest a fix.