The first build is not supposed to be your game. It is supposed to be the smallest thing that proves the core loop is worth expanding: something you can move, something that pushes back, a reason to keep playing, and a way to lose. Everything after that is refinement on a foundation that already runs.

That is also the fastest path to a useful result. A first message describing eight systems produces a project where none of them are finished; a first message describing one mechanic produces a project you can actually play, and then judge.

What happens after you send the first message

1. 1

   Oryn asks a few planning questions

   Before writing anything, it resolves the ambiguities in your description — perspective, control scheme, art direction, scope. One question at a time, each with suggested answers you can pick or overwrite with your own.

2. 2

   It proposes a plan

   You get a written plan of what the first build will contain before any files are created. Nothing runs until you approve it, so this is the cheapest place to correct a misunderstanding.

3. 3

   It edits a real Godot project

   Scenes, scripts, input maps, UI, assets, and project settings — the same files you would have written by hand, in a structure Godot opens without conversion.

4. 4

   It exports a playable web build

   The browser preview is the fast feedback loop, not the final destination. It exists so you can test the change you just asked for in seconds rather than minutes.

A first message that works

Create a third-person arena game.

Player: runs and dodges, one ranged attack with a short cooldown.
Enemies: three waves that get faster, they chase the player.
Objective: survive all three waves.
Fail state: a visible health bar reaches zero, then a restart screen.
Look: clean low-poly, bright key light, no textures needed yet.

Keep the arena small enough to read at a glance.

Notice what it does not contain: no inventory, no upgrade tree, no menus beyond restart, no story. Those are all easier to add to a game that already runs than to specify into one that does not exist yet.

How to judge the result

Play it before you read the code, and check the loop rather than the polish. The first build is a good foundation if all five of these are true:

• Movement responds the way you expected, and the camera keeps the player readable.
• The core mechanic — the attack, the jump, the grab, whatever the game is about — is present and works.
• There is a reason to keep playing: a score, a wave counter, a goal.
• You can lose, and losing is visible rather than silent.
• You can restart without reloading the page.

If the loop is right and the details are wrong, you are in the good case: everything from here is iteration, and iteration is where Oryn is strongest.