---
name: deploy
description: Produce a deployment checklist plus the Helm, kubectl, and Terraform commands needed to ship a service. Use when deploying or releasing, promoting to staging or production, rolling back, or preparing a change window.
argument-hint: "<service name>"
---

> **Recommended model:** `gpt-5.6-terra` &nbsp;·&nbsp; **Agent mode:** `ask`
>
> Skills are portable across agent clients, so they carry no Copilot-specific
> `model:` field. Set the model via the picker, a custom agent, or the
> slot mapping in `.github/model-compatibility.json`.


<!--
  SLASH COMMAND: /deploy
  MODEL: gpt-5.6-terra — reliable for structured DevOps checklists and CLI commands
  AGENT: ask — produces output for you to review, does not run commands
-->

Read the active file to identify: service name, version/image tag, and target environment.
Then produce the following — fill all `<placeholders>` from the file context:

## 1. Pre-flight checklist
- [ ] All CI checks green on the release branch
- [ ] Docker image `<registry>/<service>:<version>` exists in Artifactory
- [ ] `values-<env>.yaml` reviewed — resource limits and replica count correct
- [ ] Secrets rotated if this release touches authentication or credentials
- [ ] DB migrations tested against a staging database snapshot (if applicable)
- [ ] On-call engineer notified in #deployments Slack channel

## 2. Dry run (always run this first)
```bash
helm diff upgrade <service> charts/<service> \
  --namespace <namespace> \
  --values charts/<service>/values-<env>.yaml \
  --set image.tag=<version>
```
Review the diff carefully. Stop if any `-/+` (destroy + recreate) lines appear.

## 3. Live deploy
```bash
helm upgrade --install <service> charts/<service> \
  --namespace <namespace> \
  --values charts/<service>/values-<env>.yaml \
  --set image.tag=<version> \
  --atomic \
  --timeout 5m \
  --history-max 5
```
`--atomic` rolls back automatically if health checks fail within 5 minutes.

## 4. Smoke test
```bash
kubectl rollout status deployment/<service> -n <namespace> --timeout=3m
kubectl logs -l app=<service> -n <namespace> --tail=50
curl -sf https://<ingress-host>/actuator/health | jq .status
```

## 5. Rollback (if needed)
```bash
helm history <service> -n <namespace>          # list revisions
helm rollback <service> 0 -n <namespace>       # 0 = previous revision
```

## 6. Post-deploy
- [ ] Error rate and p99 latency unchanged in Datadog
- [ ] Alert rules re-enabled if they were silenced
- [ ] JIRA ticket moved to Done, #releases updated