AI RTS maker for a genre with a hundred moving parts

Most genres let you build up. A platformer is fun with a character and a floor. A real-time strategy game is not fun with anything less than selection, movement orders, pathfinding, a resource economy, production and at least one opposing decision-maker. Until all of those exist, there is nothing to evaluate.

That threshold is why so few solo RTS projects reach a playable state. The first playable moment sits behind weeks of infrastructure, and motivation rarely survives the wait.

Handing the code to Director 1 changes the shape of that problem. The infrastructure is written into your Godot project and you reach the playable moment early, which is when your own opinions become useful.

Unit selection including box select and control groups. Movement orders and formation behaviour. Pathfinding across your terrain. Resource gathering, build queues, tech dependencies. Fog of war. Combat resolution with counters that mean something. An opponent that makes decisions on a timer.

You ask for each in plain language and it is written as GDScript in your project. When the counters do not create interesting decisions, you say so, and they change. Balance in an RTS is found by playing hundreds of matches, and the value of the tool is how cheap each revision is.

Strategy games are largely designed through their maps. Where the second expansion sits, how many approaches a base has, whether the high ground is worth the walk. You build maps by hand in the builder, placing terrain, water, elevation and structures directly.

That is the right way round. A generated map is symmetrical and dull or asymmetrical and unfair, and skilled players find the flaw in an afternoon. A hand-built map is the artefact your game is remembered by.

RTS is expensive to run. Two hundred units pathing at once is a real load, and it is the reason browser-based strategy tools quietly cap unit counts. Stage Engine is a desktop app for Mac and Windows with Godot bundled, so the game runs on your own hardware with your own memory and your own CPU.

You need an internet connection while building, because Director 1, asset generation and world data come over the network. You do not need one to play the game you built.

The download is free with a free plan. Indie is $10 a month and Pro is $50, buying usage rather than a different engine.

The skill that makes someone good at this is not programming. It is watching a match and noticing that both players opened the same way for the fourth time in a row. That is a design fault, it is visible to anyone who plays strategy games, and it does not require you to understand a single line of what the game is doing internally.

So the working method is: play a match, notice the thing, describe it. The build order is always the same. The second resource is never worth taking. Air units make ground units pointless after minute eight. Each of those sentences is a complete instruction here, and each produces a change you can test in the next match.

The people who make good strategy games are the people who have played a great many of them and know what a stale match feels like. That knowledge is the scarce part. The code was only ever a toll on the way to using it.

Strategy games are the genre where the interface is part of the design rather than a layer over it. Selecting a group, queueing an order, seeing at a glance which of your bases is under attack: if those are awkward, the game is unplayable no matter how good the underlying systems are.

It is also the work most likely to be postponed, because it feels like polish. It is not. Build the selection and command interface early and use it constantly, because everything you learn about the game is filtered through it.