Make an Idle Clicker Game With AI: Prompt and Balance

Make an Idle Clicker Game With AI: Prompt and Balance

Describe a counter that goes up when you click, upgrades that raise it, and generators that produce while you do nothing, and Zugo builds a playable idle clicker in about a minute for 6 credits. The first build works. The balance almost never does, and fixing it is where the 3 credit edits go.

An idle clicker has no physics, no collisions and no reflexes. It is a spreadsheet wearing a costume, and the costume is easy. The spreadsheet is what decides whether anyone plays for more than a minute.

What is an idle clicker actually made of?

Four parts. A resource that increases. A manual action that increases it directly. Purchasable things that increase it automatically. And a cost curve that makes each purchase more expensive than the last, so progress always feels slightly out of reach.

That last part is the whole genre. Remove the rising cost and the game is over in thirty clicks. Keep it and you have a loop that renews itself: you save, you buy, production jumps, the next target appears, you save again. Everything else is theme.

Which is why this genre suits a text prompt unusually well. There is nothing to draw and nothing to animate that matters. You are describing numbers and their relationships, and numbers are the easiest thing in the world to say precisely in a sentence.

What should the prompt say?

Vague economy, vague game. Name these five things:

The cost curve row is the one that decides everything. If you leave it out, the build picks something arbitrary, and you will spend several edits discovering that everything is either free by minute two or unreachable forever.

Naming the theme helps more than it does in other genres, because there is so little else on screen. "A coffee shop idle game where you brew, hire baristas and open branches" produces better copy, better button labels and better upgrade names than "an idle clicker".

How do you build one in Zugo, step by step?

1. Write the economy, then the theme. Numbers first, costume second. It reads oddly but it produces a better game.
2. Wait for the verified build. About a minute. Zugo loads and runs it in a sandbox before showing it, so a page that fails to open is not handed over as finished.
3. Play for five minutes without leaving. Idle games reveal their problems over time rather than on the first screen. You are watching for the moment where nothing is worth buying.
4. Fix one number per message. "Make the second generator produce four times what the first does." "Slow the cost growth on the click upgrade." Each edit is 3 credits, with its own sandbox check.
5. Publish in one click to your-game.zugo.run and share the link.

Why does the economy break first, and how do you fix it?

Because the two curves in an idle game grow at different speeds and it is very easy to get them out of step. Production grows with what you own. Cost grows with what you have bought. If cost grows more slowly than production, you buy everything and the game ends. If it grows faster, you stall and quit.

The symptom is easy to spot once you know the shape. Play for a few minutes: if you are buying something every few seconds without thinking, the costs are too shallow. If you sit watching a number crawl toward a price with nothing else to do, they are too steep or your generators are too weak.

Fix it from the player's side rather than the maths side. Tell the build what the experience should be: "the first generator should be affordable within about ten seconds of play", or "there should always be something purchasable within roughly thirty seconds".

Those are instructions a build can implement directly, and they are far more useful than demanding a specific multiplier you have not tested. The general approach to iterating like this is in how to edit after generation.

Does progress survive a page reload?

Ask for it, because it is not automatic and it is the difference between a demo and a game. Two things are worth requesting in the same edit: saving progress in the browser so a reload continues where the player left off, and offline earnings so returning after a few hours feels rewarding rather than punishing.

Offline progress is what makes the genre work at all. Someone closes the tab, comes back tomorrow, and finds a pile of coins waiting. That single mechanic is why these games retain players, and it is one sentence to request.

Be clear about the boundary though. Saving in the browser means saving on that browser, on that device. Progress does not follow the player to their phone, and clearing site data wipes it.

A shared account with a cloud save needs a database, which means connecting Supabase to the project, and that is a real piece of work rather than a chat message. A closer look at the no-code side of the genre is in the idle game maker guide.

What does an idle clicker cost to build?

The Free plan gives 5 credits. Pro is $25 a month for 200 credits, which works out to roughly 33 builds or 66 edits, and Business is $99 a month for 800 credits.

Idle games sit in the middle of the range. The build itself is simple, but balance is genuinely iterative, so expect more edits than a tapper and fewer than a rules-heavy card game. Ten to fifteen edits is a realistic budget for something you would leave open in a tab.

Where does a prompt run out?

Prestige layers are where most people hit the edge. Resetting progress for a permanent multiplier is describable, but a satisfying prestige economy that stays interesting across several resets is design work, and it comes together over many edits rather than in one.

Competitive features are the harder boundary. A leaderboard, guilds or trading between players all need a server holding state that players cannot edit, and anything a browser holds is something a determined player can change. Zugo can wire in a database and payments, but an idle game with real stakes on the numbers is a product with a backend, not a page.

And this remains a browser game rather than a store app with push notifications calling players back. That distribution model is a different project. The genre-by-genre view of what generates well is in make a card game with AI, where the same trade appears in a different shape.

How do you publish and share it?

Publishing is one click, and the game lives at your-game.zugo.run, playable on any device from a link. Paid plans remove the Zugo badge and let you point your own domain at the game.

Idle clickers are among the easiest things to share because there is no skill barrier: anyone can click a button, and the first upgrade sells the loop by itself. Write the economy down, build it, then spend an afternoon making the numbers feel just out of reach. Free credits to start are at zugo.dev.