David Humphrey

The `/import` function can do this in CloudFlare by reading the `Content-Type` header, then passing that to the [`mime-types` module](https://www.npmjs.com/package/mime-types): ```js const mime = require('mime-types'); mime.extension('application/json')); // Outputs: 'json' ```...

The import stuff is fairly tightly bound to what happens in CloudFlare; I'm not sure how you properly fork that into val.town. However, why don't we add a `/valtown ...`...

I suspect we'll move to a more "Assistants" style db in the near future, so I'm going to close this.

I'm adding the ability to screenshot an element in #402, partially with this in mind. Why don't we take a screenshot and `POST` that to R2 as well?

One challenge of doing this well is making it easy to quickly understand what the system prompt is doing, without having to read and understand the entire thing. I notice...

I think a cool way to do this would be to support JSON metadata that describes an assistant, which can be stored in the db, stored in R2, posted on...

In #332, I'm proposing that we support `?assistant=`, which could point at the metadata for a custom JSON assistant. We should make that work for the normal app as well...

The new `@mention` thing that ChatGPT just shipped is interesting: Imagine if you could install the following assistant and give it nickname like `@superdev`: ```json { "id": "some-unique-id", "name": "Senior...

Right, my thinking is that you start out talking to one "assistant" and 3 messages in you can `@mention` some other assistant, and it switches the system prompt and model....

@taras and I discussed this today, and I wanted to capture some notes: - We'd love to be able to extract an "Assistant" from the current chat. So maybe that...