online-play 2026-10-19 20:48 ♥ 0
Really good game. I had great time learning how to make these sandwiches haha
Foon → Ludum Dare Explorer → LD60 → STACK/86
By duolapis
🤖 AI-Assisted
| Percentile | ||||||
|---|---|---|---|---|---|---|
| Category | Rank | Score | Count | LD60 | Ever | |
| Overall | 10 | |||||
| Fun | 10 | |||||
| Innovation | 10 | |||||
| Theme | 10 | |||||
| Humor | 9 | |||||
| Mood | 10 | |||||
Really good game. I had great time learning how to make these sandwiches haha
Honestly, this looks and feels like pure "vibe-coding"—is there even any part of this that was actually done by a human?
Sorry, I don’t mean to offend you; I use "vibe-coding" myself, which is exactly why I’m so familiar with the kind of interfaces that neural networks churn out.
Thanks for your comment. @voxivale In fact, the human-made parts of this game include: the design of the game’s main gameplay and terminal control scheme (in fact, this started from a paper prototype, and I kept steering the AI back from a bracket-hell kind of language to what I consider a more retro control scheme); the balancing and adjustment of level diversity; the story content and the corresponding Chinese text (though there is actually basically no story).
If you care about my creative process (or my excuses): I put together a three-person team and hoped to take part in this game jam by hand, but because this theme was hard for us to handle, we wasted a whole day trying to think of a good core gameplay and wanted to give up on this jam. But on the morning of the second day, I came up with this idea and told my teammates, but they had indeed already given up on this jam. After trying to write this core loop, I realized that the time I had left was not enough to finish the work before Sunday ended, and I had a full day of work on Monday, so most of the content of this game was made using Codex. I did spend my spare time over these two days fixing various issues, testing it myself and organizing friends to playtest, manually adjusting the level design and text, and avoiding this game become total 'slop', but it really didn’t feel enough like making a game, so I will join this weekend’s 20s game jam and make a purely human game.
By the way, I’m very much looking forward to your sharp critique or suggestions on the above human-made parts. After all, comments on the AI-implemented parts won’t help me make a better game, but challenges to my gameplay design will give me more room for reflection and improvement.
@duolapis
Regarding the game design itself, it feels quite overloaded. I think the sheer volume of text actually detracts from the game; it’s lacking in visuals—even with the help of AI, you could have incorporated some basic graphics or icons.
To start, I’d suggest cutting down on the text—perhaps keep only what’s on the left side. For the interactive buttons, you could create simple pixel-art assets; for instance, while the "Codex" is editing the code, the second window could display some basic pixel art.
That’s just my take—please don’t take it the wrong way; I was genuinely curious about how things ended up this way. Thank you for sharing your story; it’s truly saddening to hear how you were treated...
This is exactly the problem I’ve been struggling with! @voxivale In fact, some of the friends who took part in playtesting also questioned the intuitiveness of these controls, especially my artist friend. Perhaps I focused too much on the retro feel brought by the keyboard controls and plain-text display (even the awkwardness and user-unfriendly UI design is part of retro), while neglecting the players’ experience. Before submitting, the approach I thought of was to remake it in 3D, so that players could actually observe how this machine operates and what to do next, instead of feel like playing a typing game, but I didn’t have time to complete this idea. I don’t want to modify my work during this period, but if I still think this idea is good enough after the jam ends, I may optimize it this way.