deploy-prod

name	deploy-prod
description	Use when the user wants to deploy to production, "subir a prod", "desplegar en producción", "lanzar a prod", "manual deploy", "deploy prod", or promote a specific version/tag to the production environment. Use for any Afianza service. SKIP if the user only wants to release/tag (use the `release` skill instead).
allowed-tools	Bash AskUserQuestion
metadata	{"argument-hint":"[image_tag]"}

Afianza — Deploy to Production

Triggers the manual-deploy-on-prod.yml GitHub Actions workflow for the current service with a specific image tag.

Step 1 — Identify the service

git rev-parse --show-toplevel
cat package.json | grep '"name"'

Confirm with the user which repo/service is being deployed.

Step 2 — Choose the image tag

If $ARGUMENTS contains a tag (e.g., v0.11.4), use it directly and skip to Step 3.

Otherwise, show the last 5 tags:

git tag --sort=-version:refname | grep '^v' | head -5

Ask:

"¿Qué versión quieres desplegar en prod? (últimas versiones arriba)"

The image_tag input to the workflow is just the version tag (e.g., v0.11.4), not the full image URL — the workflow constructs the full registry path internally.

Step 3 — Confirm before deploying

Always confirm before triggering prod:

"Vas a desplegar <tag> en producción para <service-name>. ¿Confirmas?"

Do not proceed without an explicit yes.

Step 4 — Trigger the workflow

gh workflow run manual-deploy-on-prod.yml \
  --field image_tag=<tag>

If the command fails because the workflow isn't found, check the exact filename:

ls .github/workflows/

Step 5 — Show the run link

# Wait a moment for the run to register, then get its URL
sleep 3
gh run list --workflow=manual-deploy-on-prod.yml --limit=1

Show the direct link to the run:

Desplegando <tag> en producción. Sigue el progreso:
https://github.com/<owner>/<repo>/actions/workflows/manual-deploy-on-prod.yml

Remind the user: "Recibirás una notificación de Slack cuando el deploy termine (éxito o fallo)."