At first I didn't really want to make a full-fledged game. I just wanted to experiment with the Ren'Py engine and write a short, linear story in the vein of Southern Scrimshaw.

But then, scope creep happened.

In the first iteration of the project, the player had to solve a series of moral dilemmas. These trolley-problems-like situations would become more and more complex until a final reveal : the player is actually an AI trapped in a computer, condemned to be the moral-chatgpt of humanity, forever. 

But while kinda fun (according to the one person who played it : me), it lacked a little something.

A common challenge for visual novel creators is finding ways to add gameplay into a narrative-heavy experience.

Many add mini-games like cooking, simon says, inventory management, etc (plenty of good examples here). Others don't really mind, and it's okay, as the visual novel genre doesn't need gameplay. Choices and good writing are more than enough.

But... I must admit I'm not a big consumer of visual novels, so I wanted to have a gameplay loop, somewhere. It could underscore the character's awful situation: something so repetitive it would start to feel like —gasp—a job.

The clicker option felt obvious and I implemented it very fast thanks to Gemini 2.5 pro. So far, vibe coding has been a great method to prototype new ideas in dozens of minutes, and it helped me to test ideas and discard the bad ones without suffering from sunk cost fallacy.

Here's what it looked like :

At that point I had changed the story : you're no longer an AI answering moral conundrums but a human trapped in purgatory, answering the prayers of 8 billion anxious humans. To reach salvation, you have to click your way through all 8 billion of them in 30 minutes.

Maybe you'll have to automate that a bit, and that's where the idle clicker mechanics come into play.

But idle clicker games are usually… idle. You play a bit, spend your points on upgrades and go back hours later to check the result. My game has to be 100% active, especially as I want it to be a very short experience. So I added a gambling system.

Some prayers are worth a lot of points and can get you closer to your goal. But beware, there's a risk of losing a lot of points if you're unlucky. You basically take a bet and at first that system felt good. It adds tension between the prayers that you shouldn't answer because they are morally questionable, and the desire to gamble and win more.

But somehow it was still a bit boring ! So I added a rhythm system.

Now you get a bonus when you answer a prayer with the correct timing.

So that’s where I’m at, right now. The gameplay loop feels more interesting, and the monkey-part of my brain wants to click at the right time, often overlooking the actual content of the prayer itself. Which can lead to problems. This is exactly the "Papers, Please" style of conundrum I was looking for so I’m happy.

At least for now. 

Maybe I’ll realize that there’s a much better idea to explore, and that’s okay.

If you don’t count the weird “casino” game I did in the R coding language 10 years ago, it’s actually my first game so it’s also a very fun way to learn and explore ideas.

tl:dr : I’m having a blast.