-
Notifications
You must be signed in to change notification settings - Fork 154
Description
This project is too important to get left behind, yet the effort to bring it up to the latest 2020-12 spec seems fairly large, and there seems to be little interest in tackling it. (The $defs shim was just a bandaid.)
I'm wondering if there is any interest in enlisting the help of an LLM (Claude Code or Codex, for example). The advances in the utility of coding assistant LLMs in the last several weeks are astonishing. I got codex to estimate the effort and it came back with a concrete plan within a minute or so. (It gave the most complete coverage plan an estimate of 8-12 weeks, but that's for a human team. My guess is if the LLM did it, it would be code complete within an afternoon.)
I would be happy to dedicate LLM time to this, if there is interest. The challenge is, I am not qualified to review any code it comes up with. My knowledge of JSON schemas is very limited.