I've been working on a coding stack for AI that creates specific quality targets tailored to Home Assistant. The idea is to give the coding tool a set of quality rules and parameters to code with, as well as a validation stack to test against. The loop then will run against a given problem and achieve strong results.
With AI there is absolutely no reason that anyone using AI shouldn't be targeting "Home Assistant Platinum" tier from day 1.
This version will help you achieve Home Assistant Platinum standards, creates an audit checklist against all Home Assistant Architectural Decisions Records, pushes ruff to some stricter complexity targets (among many other ruff improvements), and gets the integration ready for HACS.
I am hoping that others share what they're doing or comment on this approach. How can I improve it? Are you doing something similar? Please share.