Skip to content
Decision guide

Before launch, confirm the deploy path is recoverable.

This deployment checklist covers the critical checks from local build to production verification. It is not a generic DevOps list: it helps you choose static publishing, VPS/CLI, GitHub Action, AI agent, or self-hosted control plane paths.

Static vs VPSCLI vs GitHub ActionAI agent deploy

Checklist

Build output

Confirm build command, output directory, and entry file. Do not treat a local dev server as production output.

Runtime config

List environment variables, secrets, ports, start command, and runtime dependencies.

Domain and TLS

Confirm temporary URL, custom domain, TLS ownership, and DNS change window.

Health and logs

Prepare health endpoint, recent logs, failure diagnostics, and status entrypoints.

Rollback

Confirm you can redeploy the previous version, or at least name recovery commands and manual gaps.

Cleanup

Preview environments, temporary resources, and failed deploys need cleanup paths.

Identify project shape first

Static sites and long-running services need different deploy paths. Identify build output, runtime, ports, environment variables, and state dependencies before choosing an entrypoint.

Credentials do not belong in code

SSH keys, provider tokens, database URLs, and API keys should live in local profiles, Appaloft secrets, or GitHub Secrets, not in the repository.

Deploy success is not verification

A live URL is only step one. Check logs, health, deep links, static assets, error pages, and rollback candidates.

Rollout steps

1

Identify deploy type

Static output, service, worker, Compose app, PR preview, or self-hosted control plane.

2

Choose entrypoint

Use CLI locally, GitHub Action for CI, Appaloft skill for AI workflows, and Cloud upload for static sites.

3

Verify recovery

Every launch should expose status, logs, health, rollback, and cleanup.

FAQ

What projects is this checklist for?

Static sites, VPS services, GitHub Actions, AI agent deploys, and self-hosted control plane deployments.

What should I check after deploy succeeds?

At minimum: URL, critical pages, static assets, health, recent logs, error behavior, and rollback path.

When should I avoid static publishing?

When the project needs API routes, SSR, workers, cron jobs, or a long-running Node/container process.

Related topics

Keep following the same deploy path.

Appaloft SEO pages are organized around real deployment tasks. Each page should lead to the next useful step, not stand alone.