what mad science is this

Awesome

Wow, this is seriously cool! It's amazing that you're using ContextVM for this. I remember the first time I saw that powerful addone tool 🥹 I have a couple of questions. Is it true that the JSON object of a hypernote is agnostic to the code and the libraries rendering it? Is there a specification with guidelines on how to translate hypernote syntax? Also, I noticed that for tool usage, there is a field 'tool_call' set to true. Is that because this action is also responsible for handling the response? Anyway, awesome! 🔥🔥

> Ideally this will soon be in a good way to make fun little UIs for any contextvm tools out there! Yes, that sounds amazing! You could also take the inverted approach: from a contextVM server, see what hypernotes use it. I think it's going to be a lot of fun exploring these ideas. Regarding the wire up, do you already have a specification for the HN JSON output? Additionally, do you think it would be worthwhile to add a registry of supported UI elements? This could help avoid naming convention issues and assist other developers in adhering to a common set of UI elements