The Orbit flywheel
Good software is never one prompt.
A prompt gives you a demo. What turns it into a product you can ship to production is the work that comes after, one turn at a time. Orbit is built for that loop, and makes every turn effortless.
One idea, all the way round.
Each turn works on the same app, so nothing you build gets thrown away.
Explore the idea before you build it.
An idea rarely arrives whole. Explore is where you get to know it: gather the notes and links, ask the open questions, and let Orbit research the ones you can't answer yet. The scratchpad keeps it in one place. You build the app later, once you understand what you're building.
Pitch it, then scope it.
A researched idea is still a guess about what to build. Shaping turns it into a structured pitch, then a scope of work you can read: what gets built, in what order, and what is still open. You settle it while changing your mind is free.
Build from the data up.
This is the hard part, and the one that matters. A product is built up in layers: the data model that holds it, the workflows that carry your business logic, and the UI that presents it to the people using it. A demo fakes all three. In Orbit each one is a surface of its own, over a single data model.
Add a field once, and it is there for every screen and workflow to use.
Production is a turn, not a finish line.
Going live is where the app meets people it never planned for: strangers signing in, saving their work, finding the corners you never tested. Orbit puts it online with a database and automatic HTTPS, plus your own domain and sign-in when you want them. Not the version that looks done, the version that stays standing when people show up.
Let what you learn start the next turn.
The live app teaches you what no plan could, and those findings start the next turn. You can even point an agent at the running app to make the change. This time the domain is yours, the workflows run, the screens know what a customer is. That is the difference between a loop and a flywheel: the wheel keeps the work you already did.
Round again.
Grow feeds Explore, and the wheel keeps the work you already did.
The first turn
One prompt starts the flywheel.
“A CRM for investor outreach.” From that one prompt, Orbit builds a first version, live on a URL. Not a demo that breaks. Not the finished product either. A starting point you take around the loop.
You don't need it all figured out.
You have an idea and a stack of questions. That is exactly where the loop starts. Describe it, and Orbit builds the first version, then takes each turn with you, until it is an app people use.