How well do LLMs perform at generating Elysia code?

[hughcrt]

Since this is a relatively new framework, I was wondering about your experience using Codex or Claude for generating code in Elysia projects. I’m a bit concerned about hallucinations, especially given the models’ knowledge cutoffs, compared to more established frameworks like Express.