Lawn Mower Checker
"simultaneously ridiculous and very cool"
That was the reply by someone when I shared a recent build.
This isn't a blog that's going to walk you through a "game changer" or other business cliché, but hopefully you take away a simple idea.
That is --> have fun as you develop your technical skills.
Negativity is abundant across all forms of media and it is a poor representation of how work should and can be done. You might need to hear this, but it's okay to have fun with work, it's okay to build something that will never be monetized, and it's okay to build simply to build because that's what you enjoy doing.
On to the build.
The Lawn Mower Checker
So there I was. Monday evening and I knew I wanted to experiment with AI.
I get bored easily and many of the AI tutorials I see are bland with too much "optimization," i.e., the person doing the training is trying to sell how you can optimize product X.
I didn't want that.
Instead, I set a "test" for myself.
I gave myself an hour to:
Find an open API
Use Postman to get JSON data from the API
Pass the info to Claude Code
Connect Claude Code to my GitHub
Build a retro-looking page and get it live
These are the things I do for fun, and in practice, they help with work because the learnings transfer to work projects. Of course, the lawn mower checker isn't used in work, it's more about identifying better ways to connect data, a small learning in Claude Code, and many times I simply stumble onto a feature in an application that I have never seen before.
Back to the test.
I did a basic Google search and was overwhelmed by the number of open APIs. That made me stop and think about things I like.
As a full-time dad, of course I thought about the weather, which led me to think about when I will mow my lawn on Saturday.
That got me on a path and into a brainstorming session with Claude (chat) on what could be built. We went back and forth and thought of a simple landing page that looks at the heat index for my town and precipitation. Then it recommends when to mow or not to mow (i.e., mow early if it's hot or don't mow if it's raining).
That's it.
No monetization, no hustle and grind, no optimization. Just something silly for the sake of building.
Once the plans were locked in, it was relatively straightforward.
Claude helped me write a URL for Postman that pulled JSON data to verify the API was working
I used Claude Code in my local terminal to do a little back and forth building out the code for the page, inclusive of the API call
Claude Code was already connected to my GitHub, so all I needed to do was point Claude Code to the repo I built - I created the repo in parallel as Claude was coding
We pushed it to GitHub and enabled GitHub pages
Then we did what I consider fun, a little back and forth on aesthetics, what to show, font sizing, adding a favicon, etc.
It took a little less than an hour, and I stopped the session with a big smile. There were a few small things I had not done before, mainly the full end-to-end build with the open API. That gave me some confidence as I seem to be using APIs more frequently at work. This fun session helped me troubleshoot some items that, coincidentally, came up later in the week.
I showed a few people and got a similar reaction, with the best being that the tracker is "simultaneously ridiculous and very cool."
Since you are dying to see the checker, here you go --> https://txcountrynerd.github.io/yard-planner/
