That's a different problem. The benefit of stacked PRs is that they let
you review each piece on its own, which helps you/team keep mental alignment
with changes in the codebase and disregard individual slop layers.
I had this idea to build a webpage that can rewrite itself on the edge using a local model and a simple harness in the frontend, while playing around with storing HTML state in the URL.
Currently supports Gemini Nano (local), Anthropic, or OpenAI.
Gemini Nano is a lot less impressive than the SOTA models, but it's still fun to see it work!
reply