Because it's complex and I'm stuffing 30-50 PRs a day through it. 7 GHA workflows across four self-hosted runners.
I'm on my third iteration, after constantly log jamming previous versions. Most steps aren't "run pytest", they're gates that guard against an LLM's bias to continuously add more and more complexity to a system.
I'm aware of the irony of having a complex system to mitigate complexity, but the key difference is these rules bound complexity growth. If you're legitimately interested in the details, let me know. I'm too tired to write up much more but would be willing to drop in a LLM-authored summary of the details.
What are the project's capabilities? I have also written a personal finance application, and it's ~3k LOC total. And have several projects that are dramatically more complicated than a typical personal finance app, and are 30-50k.
This begs the question: If you're using LLMs heavily, wouldn't it be better from a time, correctness, and token perspective to have a more compact and maintainable code base? We don't know what the personal finance app does, how quickly it runs, how much it improves your finance management, how easy it is to install and update etc, or how maintainable it is by people or LLMs. What you did post is a red flag of consequence, without mitigating factors: We know only the red flag.
Another perspective: I use LLMs heavily as well, but don't find they appreciably increase code verbosity, for a given set of features/requirements.
I work for a gigantic financial institution with ~2000 devs pushing code daily with 100s of apps that need to be built and I'd be shocked if we were above 50kLOC for all the CI/CD stuff. It's probably far less considering the point of CI/CD is reproducibility and ease of adding to...
