deploy

name	deploy
description	Deploy the application to staging or production. Validates, builds, pushes Docker image, deploys via Helm, and runs smoke tests.
argument-hint	<staging|production> [--skip-build] [--dry-run]
disable-model-invocation	true
allowed-tools	Read, Bash, Grep, Glob, Edit

Deploy Application

Deploy this application to the $ARGUMENTS environment.

Parse the arguments:

• First argument: environment (staging or production). Default to staging if not specified.
• --skip-build: Skip Docker build/push, deploy with existing image
• --dry-run: Render Helm templates and show diff without actually deploying

Pre-flight Checks

Before deploying, verify ALL of these:

1. Git status is clean - no uncommitted changes (warn if dirty, ask to continue)
2. On correct branch - develop for staging, main for production
3. Read package.json to get the current app version
4. Sync Chart.yaml appVersion with the version from package.json
5. Helm lint passes with the target values file:

   helm lint helm/goravel-blog -f helm/goravel-blog/values.<env>.yaml
   

Build Phase (skip if --skip-build)

Read DOCKER_REGISTRY and DOCKER_IMAGE_NAME from .env (fallbacks: docker.io and books-database). Construct the full image name: $DOCKER_REGISTRY/<dockerhub-username>/$DOCKER_IMAGE_NAME

1. Build Docker image:

   docker build -t $DOCKER_REGISTRY/<username>/$DOCKER_IMAGE_NAME:<tag> \
     --build-arg BUILD_DATE=$(date -u +'%Y-%m-%dT%H:%M:%SZ') \
     --build-arg VCS_REF=$(git rev-parse --short HEAD) \
     .
   

   • For staging: tag = develop
   • For production: tag = version from package.json

2. Scan image with Trivy (if available):

   docker run --rm aquasec/trivy image --severity CRITICAL,HIGH $IMAGE:<tag>
   

   Report findings but don't block unless CRITICAL vulnerabilities found.

3. Push image:

   docker push $IMAGE:<tag>
   

Deploy Phase

Dry Run Mode (--dry-run)

Render templates and show what would change:

helm template goravel-blog helm/goravel-blog -f helm/goravel-blog/values.<env>.yaml
helm diff upgrade goravel-blog helm/goravel-blog -f helm/goravel-blog/values.<env>.yaml -n <namespace>

Stop here for dry-run.

Actual Deployment

1. Determine namespace: books-staging or books-production

2. Run Helm upgrade:

   helm upgrade --install goravel-blog helm/goravel-blog \
     -f helm/goravel-blog/values.<env>.yaml \
     -n <namespace> --create-namespace \
     --set image.tag=<tag> \
     --atomic --timeout 10m \
     --wait
   

   For production, ask for explicit confirmation before executing.

3. Verify rollout:

   kubectl rollout status deployment/goravel-blog -n <namespace> --timeout=300s
   

Post-Deploy Verification

1. Check pod status:

   kubectl get pods -n <namespace> -l app.kubernetes.io/name=goravel-blog
   

2. Get ingress URL:

   kubectl get ingress -n <namespace>
   

3. Smoke test the health endpoint (if accessible):

   curl -sf https://<host>/health
   

Rollback

If deployment fails, Helm --atomic auto-rolls back. If post-deploy checks fail, offer to run:

helm rollback goravel-blog -n <namespace>

Summary

Output a deployment summary table:

Item	Value
Environment	staging/production
Version	x.y.z
Image	$DOCKER_REGISTRY//$DOCKER_IMAGE_NAME:tag
Namespace	books-xxx
Status	SUCCESS/FAILED
URL	https://...