Why I'm building 30 apps in 30 days

I keep hearing that agents are going to change how software gets built. The only way I know how to find out whether that's true is to build things — a lot of things, quickly, and pay attention to what actually changes in the process.

So here's the dare I made myself: 30 useful apps in 30 days. Not toy demos. Things I, or someone I know, would actually open twice.

The rules

• One app per day, shipped somewhere I can show it.
• Agents do as much of the work as I can responsibly delegate.
• I write down what I learned while it's still fresh.

What I'm actually trying to learn

The apps are the excuse. The real goal is to get fluent in agentic software engineering: how to scope work for an agent, where they shine, where they quietly go off the rails, and how the shape of a product changes when the cost of a first version drops to almost nothing.

This site is the logbook. The apps ledger tracks everything I build, with an honest verdict on how each one turned out. The journal is where I think out loud.

Let's see how far 30 days takes us.