An AI RPG maker that leaves the engine in place

A lot of people searching for an RPG maker have used one of the classic tile-based editors and liked it right up to the point where they wanted something the editor did not offer. Those tools are genuinely good at what they cover. Their limit is that the coverage is a list, and the moment your idea sits outside the list you are writing plugin code in a scripting language you did not sign up to learn.

Flockbay is built the other way round. There is no list of supported RPG features, because the layer that implements features is Director 1 writing GDScript rather than a menu someone shipped last year. If your combat needs a mechanic nobody has built before, that is a normal request rather than a feature gap.

The trade is that this is a download, not a browser tab, and the engine underneath is Godot rather than a bespoke runtime. Your project is a Godot project. It opens in Godot. That is deliberate, and it is the part that keeps the ceiling off.

An RPG is mostly bookkeeping wearing a costume. Stats that change, items that move between containers, dialogue that branches on a flag you set two hours ago, quests that track state across saves, enemies that scale to something. None of that is glamorous and all of it has to be right, because a save file that loses your inventory ends the game for that player.

That is exactly the work to hand to Director 1. You describe the rule in plain language: levelling curve, how resistances apply, what happens when a quest item is sold by accident. Director 1 writes the code into your project, and you play it to find out whether the rule you described is the rule you wanted.

You will find that it often is not, and that discovering it takes ninety seconds of play rather than an afternoon of reading. Not writing the code does not mean not judging it. The judging is your job and it is the part that decides whether the game is any good.

The map, the towns, the dungeon that the second act hinges on: you build those by hand in the builder, because describing a village in a sentence is slower and worse than putting one there.

Which kind of world you are putting it in is the first question the app asks — 2D or 3D, shown as two pictures, with no default and nothing created until you pick. On the 3D side you place terrain, water, weather and buildings directly, the way you would in a level editor.

In a flat world you paint the ground and place the town. Drag to lay ground and drag to erase it, and the tiles, the edges and the collision all fall out of which cells are ground, so the road you dragged is a road the party walks along. The houses, the stalls and the well are pictures put at an exact position, turn and size, each sitting on its own drawn bottom-centre, so a house stands on the ground and whatever is lower down the screen draws in front.

This is the half of RPG development where taste is the whole skill. A corridor that reads as tense, a town square that makes you want to talk to someone, a view that earns the walk up the hill. Nobody can specify that in a prompt, and pretending otherwise produces the flat, sample-scene look that gives generated games away instantly.

Asset generation is available while you build, and so is anything you bring yourself. Building needs an internet connection because generation and world data come over the network. Playing the game you built does not.

RPGs are the genre where scope kills projects. Forty hours of content is forty hours of content whether or not the code was written for you, and the writing, the quest design and the pacing are still yours. Flockbay removes the programming barrier. It does not remove the work.

What it does change is the cost of trying things. A combat system you can rewrite in an evening is a combat system you will actually iterate on, and an RPG that has been iterated on is the difference between a demo and a game people finish.

Start smaller than you want to. One town, one dungeon, one loop that feels good to repeat. If that is fun, the rest is production.

The temptation with an RPG is to ask for everything at once: combat, inventory, quests, dialogue and saves in a single request. It will produce something, and you will have no idea which part is responsible when it does not feel right.

Adding one system, playing it, and only then adding the next is slower for about an hour and faster for the rest of the project. It also means each piece is judged on its own, which is the only way you find out that the levelling curve was the problem all along.