The demo worked, but could I actually maintain the app?
by•
Building a working prototype with AI can feel like the hardest part is over. Then you come back a week later to change one small thing, and suddenly you’re dealing with unfamiliar code, broken dependencies, or a feature that affects three other parts of the app.
I’m starting to think the real test of an AI-built app isn’t whether it works on day one, but whether you can understand and maintain it on day thirty.
For those building with AI, what do you do to make sure you actually understand what you’ve shipped?
7 views
Replies
Two things carry most of the weight for me. First, one shape for everything. Every screen in my app is the same pair, a view and a view model, every feature has the same folders, and the whole layout was copied from an older project I already knew. When the AI writes a new screen it lands in a place I could find blindfolded, so day thirty is reading, not exploring. Second, the repo keeps its own memory. There is a notes folder with one file per area, payments, auth, the music engines, and every surprise goes into the matching file the moment it is found, with the reason. Before touching an area, the AI and I both read that file. The code tells you what it does, the notes tell you why it is weird, and the why is what you forget first.