In our house, Everyone Watches Women’s Sports. Everyone also loves trivia and knowledge games. So when we found Immaculate Footy, we fell in love. And immediately, we lost interest. Men’s soccer is great, sure, but we are NWSL fans through and through. Rather than giving up, I gave Claude a chance to build a version of Immaculate Footy specific to the NWSL.
So last night, I fired up Claude Code and typed up the basics of what we were looking for:
Rebuild "Immaculate Footy", but scope this version specifically to the NWSL. * Goal: Design a single-page app to mirror the functionality of Immaculate Footy * The y-axis will be NWSL teams * The x-axis will be any of the various NWSL statistics tracked on https://fbref.com/en/ (or other NWSL stats page, if that one's unavailable). * I'm open to the y-axis also including women's national teams, but all answers must be NWSL players. Your output should include: * A prototype showing the look, feel, and user interactions of the page. Minimalist design. * A connection to a stats database like fbref.com or similar. * A list of y-axis selections. * A list of x-axis selections. My eventual aim is to host this as a single page app on my squarespace site, stevecahill.ai. Let's get to work toward that end goal.
I woke up this morning to a Claude notification letting me know that an offline prototype was available. I played the game straight from the Claude app on my phone. Its super fun, but not perfect. I gave Claude a few pieces of feedback and instructed it to maintain a backlog of capabilities to add later. I published the prototype, and here it is:
This took me:
5 minutes to craft the Claude prompt
20 minutes for Claude to build (Fable 5.1 on ultracode - cost $1.52 in usage credits)
5 minutes to publish the GitHub Page
This is clearly a prototype, but its an awesome demonstration of how rapidly we can now convert from idea to execution. I didn’t write a requirements document. No list of user stories. I didn’t need to create wireframes. I didn’t need to spend days writing JavaScript to handle edge cases. In the “before years”, this all would have taken me roughly a week of elapsed time, maybe more, to code up by hand…with multiple late night sessions locked to my desk, staring at my computer screen.
But instead, I wrote the prompt, took my dog for a walk, and came back to a working prototype. Some people call this a “productivity multiplier”. But not me. I think of AI as a “life multiplier”. I can deliver raw ideas, straight from my head to prototype, in a single morning. And I can spend the reclaimed time with my family, with friends and colleagues, building human-to-human connections away from the screens that have demanded so much from us for the past 20 years.
Speaking of which…I’ve got Claude working on some other ideas. I’ll share those soon. Time for me to log off, walk the dog, and enjoy the crisp autumn day.
