The real shift happened when I started integrating an AI workflow into my development process. I used myCat as a sandbox for AI experiments—first with ChatGPT and then moving into Cursor—before I felt comfortable applying those tools to paid client work. Using an LLM agent to refactor my messy early logic helped me realize that "perfect" code is a myth and that the value lies in the functionality and the iteration.
I've finally moved the project to open source. It's a simple screen-pet program, but it represents my transition from a timid beginner to a dev who embraces AI-assisted coding.
For anyone interested in seeing a real-world example of a small-scale Python project or wanting to see how a beginner's code evolves with AI help, the source is available here:
https://github.com/yumiaura/myCat
If you're a seasoned dev, I'd appreciate any technical feedback on the architecture. For those just starting out, I highly recommend finding a "pet project" like this—it's the only way to make the syntax actually stick.
Next NovaOS: Building a 16-bit OS from Scratch →