Regex Blaster(mdp.github.io)
mdp.github.io
Regex Blaster
https://mdp.github.io/regex-blaster/
57 comments
Every vibe coded site is too dark and the text is too small.
They all have this rounded box design as well. I wonder where that came from, I don't think it was a predominant style before.
Recently asked Codex (GPT-5.2) to write a small single-page HTML frontend to debug some REST endpoints. As it was just a one-off tool, I put in no instructions about looks or styling at all. Lo and behold, the tool it wrote came with exactly that round-box style.
It seems to be the "default" style of some models for some reason.
Which makes me wonder if people already experimented with different style suggestions to get different results: "Make it look like an 1998 GeoCities page" / 2005 Facebook / Newgrounds / DeviantArt / HN / one of those Windows XP simulators with built-in window manager / etc
It seems to be the "default" style of some models for some reason.
Which makes me wonder if people already experimented with different style suggestions to get different results: "Make it look like an 1998 GeoCities page" / 2005 Facebook / Newgrounds / DeviantArt / HN / one of those Windows XP simulators with built-in window manager / etc
I vibe code web apps with Google's Gemini and I think it actually mimics Google's UI and UX because I see similarities between my vibe coded web apps and Google's web apps.
But that's a different style from the these colorful border rounded boxes that I think Claude in particular loves to produce.
Every vibecoded site have this same dark look with shining hue-gradient borders, can't wait for the future the entire web be filled with this generic look
And not playtested at all :D
This is fair, although I ask for it to be dark themed to match what I think was the style of typing game I remember growing up with (it's been a while). Bumped up the font though.
Next time please ask it to respect system dark/light mode preference, it's trivial to do, especially for an LLM which can spin up light/dark alternatives easily.
no
considering free windows being light theme only, it should be a button, not a "system default"
considering free windows being light theme only, it should be a button, not a "system default"
By "free windows" do you just mean an unactivated copy of Windows? That doesn't prevent the user from configuring their preference in the browser itself.
There should be a button too, but it's simple to add a line so that it also defaults to any provided preference.
That's fine, too. Either way, give the user the choice.
My top complaint is that if I've successfully used a pattern, I want my text removed. I keep forgetting to backspace a bunch, then get frustrated that my pattern isn't working.
Other than that, great game!
Other than that, great game!
And all the text is grey-on-grey and basically unreadable. Not to even mention accessibility.
Automated accessibility testing needs to be in your loop, whether you are using an llm or not. Aria labels are easy to get right but they are also easy to forget.
I could envision the style even before clicking on the site.
Maybe because it 1337 hackerman-style, or something.
What evidence do you have that this is vibe coded?
Nice game!
We made a similar game several years ago for the Pyweek game competition, but there wasn't the fun "letter invaders" style that this one has.
https://pyweek.org/e/RegExExpress/
I really like your implementation!
Might be good to limit some of the special operators to give more focus -- otherwise the early levels are a bit too solvable with ".*"
We made a similar game several years ago for the Pyweek game competition, but there wasn't the fun "letter invaders" style that this one has.
https://pyweek.org/e/RegExExpress/
I really like your implementation!
Might be good to limit some of the special operators to give more focus -- otherwise the early levels are a bit too solvable with ".*"
I don't understand the first "combat" level. There's no real defining pattern separating the good from bad hex strings, so it's just a typing speed contest to type all the enemy patterns, right? What am I missing?
As far as I can tell, the first combat level enemies all start with "ALERT-" and have exactly 3 digits.
sorry, second combat level. It's all 6 digit hex color strings, some good some bad.
Ah, I see. Yeah, that one definitely took me multiple attempts to see what it wanted.
I believe that the "enemies":
1. Must start with "#"
2. Must be exactly 6 hex digits
3. Must be lowercase
I believe that the "enemies":
1. Must start with "#"
2. Must be exactly 6 hex digits
3. Must be lowercase
I cant even read this because most of the text is outside my phones viewport. Please test your stuff before posting it here.
Wow really cool! Genuinely fun, and educational at the same time.
One usability request: after firing a regex, could the text box be cleared? It's not hard to hit Ctrl+A and start typing again, but it does add a bit of friction. (I can send a PR)
One usability request: after firing a regex, could the text box be cleared? It's not hard to hit Ctrl+A and start typing again, but it does add a bit of friction. (I can send a PR)
I played through the training level and it took me a minute to realize it's sub string/grep match. So in the first training level the pattern . matches everything when it feels like it should be .* to match all the characters of the enemies.
The UI looks great!
The UI looks great!
Source code!: https://github.com/mdp/regex-blaster
Really fun. I just wish that stats were saved in localhost (like combat mode progress) and that esc instantly ended the round if you know you will lose.
Cute. I had a crack at it but for me the fatal flaw is having to specifically delete the last regex before entering the next one, especially as the action gets tight. Each regex should go away once it's submitted, duh...
Haha, this is nice. I'm bad at regex most of the time. Playing this felt like when I first switched from Visual Studio to Vim. it’s a bit of a learning curve. It’s an interactive game btw
This is really funny ;D Gives Tetris vibes and is executed beautifully.
The colors are difficult for colorblind people. Orange/green is difficult already, and then green turns into red depending on the state? Ugh. Looks fun but unplayable for me.
The page width is not responsive and unusable on mobile
I wish it wasn't time-limited...
This looks like something I would vibe code with Google's Gemini. Interesting concept.
cool idea but it needs to get slower as the levels get harder
Claude default frontend skills suck :D
Cool idea! I shall give it a try :)
Fun interactive game!
TL;DR: I think you should still learn regex, even though AI has made it a "useless" skill
https://mdp.github.io/2026/03/17/the-kids-are-alright-and-th...
https://mdp.github.io/2026/03/17/the-kids-are-alright-and-th...
Not so useless. In my experience LLMs are about 50/50 on making a regex that actually works and covers the cases you asked it for. Even less when you get into cases needing advanced features like backreferences and lookahead.
Anecdotal data point, writing and maintaining regex is still a core part of my job. Not useless at all for me :)
A little bit early to tell.
Let’s wait how affordable, available and good AI is when the companies turn to profit maximization and enshittification begins
Let’s wait how affordable, available and good AI is when the companies turn to profit maximization and enshittification begins
You can go local now with qwen 3.5 9B Q4 powering hermes agent at 35 to 50 tok/s with 99 percent tool call success rate on a used RTX 3060 for the price of two months of ChatGPT Pro and never bother. https://xcancel.com/sudoingX/status/2033020823846674546#m
This is the worst local AI will ever be. It only gets better from here. https://xcancel.com/sudoingX/status/2033959603944493192#m
This is the worst local AI will ever be. It only gets better from here. https://xcancel.com/sudoingX/status/2033959603944493192#m
Nope, if nobody trains the models on new data you have at some point an outdated model.
Imagine Qwen 3.5 created in the 1990s and then use it for today web or desktop development.
And is the problem solved that training AI with AI code makes the AI worse? If not the "it only gets better" claim is questionable.
Imagine Qwen 3.5 created in the 1990s and then use it for today web or desktop development.
And is the problem solved that training AI with AI code makes the AI worse? If not the "it only gets better" claim is questionable.
> Nope, if nobody trains the models on new data you have at some point an outdated model.
As people train the models on new data they'll be increasingly training on AI output including hallucinations and slop. More garbage in means even more garbage out and the cycle will continue as "updated" models decline in quality.
As people train the models on new data they'll be increasingly training on AI output including hallucinations and slop. More garbage in means even more garbage out and the cycle will continue as "updated" models decline in quality.