se-k8-deploy

Deploy and undeploy the AI Demo application to the Ping SE DevOps Kubernetes cluster.

1|Updated Jun 25, 2026
One-click install
npx skills add https://github.com/curtismu7/AI-DEMO2 --skill se-k8-deploy-curtismu7
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: se-k8-deploy
Source: https://github.com/curtismu7/AI-DEMO2/tree/main/.claude/skills/se-k8-deploy
Command: npx skills add https://github.com/curtismu7/AI-DEMO2 --skill se-k8-deploy-curtismu7

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Deploying this demo to the shared Ping SE DevOps Kubernetes cluster (ping-dev-aws-us-east-2) involves many fragile steps — kubeconfig setup, OIDC login, GHCR image pushes, namespace derivation, and PingOne callback fixes — and missing any one causes confusing failures. ## Core Features & Use Cases - Guided deploy routine: Runs the canonical ./run-pingaws.sh launcher with GHCR auth refresh, Docker AI plugin workaround, and compose profiles for the full 13-image build. - Prerequisites checklist: Verifies DEVHELP namespace access, Secret Server kubeconfig, kubelogin/kubectx tooling, and the us kubectl context before deploying. - Troubleshooting table: Maps symptoms like context deadline exceeded, stuck GHCR pushes, and wrong OAuth callbacks to concrete fixes. - Use Case: A Ping SE wants to demo at https://ai-demo.ping-devops.com; the skill walks them through cluster access, image push, manifest apply, verification, and the mandatory undeploy when finished. ## Quick Start Deploy the demo to the Ping SE cluster by running ./run-pingaws.sh after confirming kubectl context us is active.

Frequently Asked Questions about se-k8-deploy

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I deploy the AI demo to the Ping SE Kubernetes cluster?▼

Run ./run-pingaws.sh from the repo root after refreshing GHCR auth with gh auth refresh -h github.com -s write:packages and confirming kubectl context us is active. The script builds and pushes 13 images to GHCR and applies the manifests.

How do I get access to the ping-dev-aws-us-east-2 cluster?▼

File a DEVHELP JIRA service request titled Grant Access to AWS GTE K8's Cluster for ping-dev-aws-us-east-2, then download the kubeconfig from Ping Secret Server into ~/.kube/config and authenticate with kubens ping-devops-<you> via PingOne browser login.

Why does kubectl fail with no context exists with the name us?▼

This error means the SE cluster kubeconfig is not installed. Download the config from Ping Secret Server, place it at ~/.kube/config with chmod 600, then run kubectl config use-context us.

Why does the GHCR push hang after docker compose build?▼

The docker-ai CLI plugin can hang after builds on some Macs. Disable it by renaming ~/.docker/cli-plugins/docker-ai to docker-ai.disabled before running the deploy, then re-enable it afterward.

Why does login redirect to localhost instead of ai-demo.ping-devops.com after deploy?▼

The PingOne application still has local callback URLs. Run ./se-update-pingone.sh or PUBLIC_APP_URL=https://ai-demo.ping-devops.com npm run pingone:bootstrap to update the redirect and post-logout URLs.

Do I need to undeploy from the shared SE cluster when finished?▼

Yes, undeploy is required because the cluster is shared and publishing rights may be revoked otherwise. Run ./run-pingaws.sh undeploy when the demo session ends.