VE

verify-prod-deployment

Health-check automation for OpenShift production clusters post-deployment.

Install

mkdir -p .claude/skills/verify-prod-deployment && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/14705" && unzip -o skill.zip -d .claude/skills/verify-prod-deployment && rm skill.zip

Installs to .claude/skills/verify-prod-deployment

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Use when the user says they did a fresh prod deployment and wants to know if everything is okay. Checks pods, deployments, and events on the jotnar-ymir OpenShift cluster.
171 chars✓ has a “when” trigger
Intermediate

Key capabilities

  • Inspect recent git logs for changes
  • Check the running status of pods in OpenShift
  • Verify deployment rollout status
  • Note updated imagestreams
  • Flag warning severity events in the cluster

How it works

This skill verifies a production deployment by checking git logs, pod statuses, deployment rollouts, imagestream updates, and recent cluster events on the jotnar-ymir OpenShift cluster.

Inputs & outputs

You give it
User confirmation of a fresh production deployment
You get back
Report on deployment health including pod status, deployment status, imagestream updates, and recent events

When to use verify-prod-deployment

  • Verifying prod rollout success
  • Debugging deployment pod restarts
  • Auditing OpenShift cluster events

About this skill

Verify Production Deployment

Overview

Check the health of the jotnar-ymir OpenShift cluster after a fresh deployment. Run these commands in order and report findings.

Steps

1. What changed

git log --oneline -25

2. Pod status

oc get pods -n jotnar-ymir--jotnar-ymir

Expected: All agent/service pods Running, completed jobs Completed, zero Error or CrashLoopBackOff or ImagePullBackOff.

3. Deployment rollout status

oc get deploy -n jotnar-ymir--jotnar-ymir

Expected: All deployments show READY count matches desired (e.g. 2/2, 1/1).

4. Imagestreams

oc get imagestream -n jotnar-ymir--jotnar-ymir

Note the UPDATED timestamp for each imagestream. In the report, call out any agent or mcp-server imagestream updated within the last 24 hours — these are the ones that just rolled out.

5. Recent events

oc get events -n jotnar-ymir--jotnar-ymir --sort-by='.lastTimestamp' | tail -40

Look for Warning severity lines — Normal is noise. Flag anything other than routine Pulling/Pulled/Created/Started/Completed.

Reporting

Report as a short table or bullet list:

  • All pods running: yes/no
  • Any restarts: yes (pod name + count) / no
  • Any warnings in events: yes (what) / no
  • Deployments all READY: yes/no
  • Imagestreams updated <24h: list each (name, tag, updated) — highlight agents and mcp-server
  • Overall verdict: OK or ISSUE: <what>

Known gotchas

  • Recreate strategy: Deployments use Recreate (not RollingUpdate). During rollout there is a brief gap with 0 pods — normal, not a failure.
  • Quota deadlock: If new pods show FailedCreate + quota error, check oc get appliedclusterresourcequotas. Usually resolves itself with Recreate strategy.
  • jira-issue-fetcher jobs: These are cronjobs that run every 5 minutes and complete quickly — Completed status is expected and healthy.

When not to use it

  • When the deployment is not on the jotnar-ymir OpenShift cluster
  • When the user is not asking about a fresh production deployment
  • When the task is to debug a specific application issue rather than overall deployment health

Limitations

  • Limited to the jotnar-ymir OpenShift cluster.
  • Focuses on overall deployment health, not specific application debugging.
  • Assumes `Recreate` strategy for deployments, allowing brief gaps with 0 pods.

How it compares

This skill provides a standardized, systematic check of multiple OpenShift components to assess overall deployment health, unlike a manual, ad-hoc inspection.

Compared to similar skills

verify-prod-deployment side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
verify-prod-deployment (this skill)01moReviewIntermediate
ops01moReviewAdvanced
deployment-pipeline-design62moReviewAdvanced
observability-engineer124moNo flagsAdvanced

Try saying

Example prompts that trigger this skill in your AI assistant.

Search skills

Search the agent skills registry