Short answer
You do not rebuild anything. Bring the repository you already have, push it, and it goes through the same pipeline as every other Forge app: dependencies resolve, checks run, and you get a live URL in about a minute.
What you bring
One thing: the code. Forge does not ask where it came from. A project you wrote over two years, a weekend build, or an app an AI assistant scaffolded last night — all of it arrives the same way, as a codebase ready to run.
What happens on the way in
- Dependencies resolve automatically, so there is no hand-built environment to set up.
- The app is checked for obvious problems — broken pages, missing assets, exposed secrets — before anyone sees it.
- You get a working URL, running close to your users, billed only while it actually handles requests.
A cutover order that avoids surprises
Deploy here alongside whatever currently serves the app. Click through the pages your users actually hit, not just the homepage. When the checks pass and the pages behave, point your users at the Forge URL. If the app is client work that needs to keep its own address, custom domains are part of the offering — the pricing page shows which tier includes them.
Notice what is missing from that list: configuration files, manual server setup, an environment to babysit. That is usually the part of a migration that eats the whole day, and here it is simply not on the agenda.
When to think twice
If your app leans on something unusual — a niche dependency, a specific background process — deploy it here first and let the checks tell you, before you cut anyone over. The pipeline is identical for everything, which is exactly what makes a dry run cheap.
The takeaway
Migration here is not a project; it is a deploy. Push the code, watch the checks pass, and switch your users over when you are satisfied with what you see.
Related reading: What Forge checks before your deploy goes live and Where does Forge host your apps?.