The landing page looks lovely, but the examples shown have no foundation in my head for what they're doing / showing / helping with.
It's confusing enough that I don't know if this is just way above my knowledge level or whether I just don't understand what it even is.
It feels like you've developed something really useful but it's so internally obvious in your own head that it's missing some steps in the explanation for everyone else on how to come along with you and see the power of it.
I'm actually using Radix right now to brainstorm and update the example on the landing page to be more descriptive.
If it helps: the basic idea is that instead of an agent returning text or markdown we get it to return HTML. We then provide an app which makes working with that HTML actually useful (i.e. a sidebar to list the tools you've built, a way of commenting on the pages it generates).
The most trivial example I can think of is creating a simple colour picker. With Radix you can get an agent to add a colour picker that updates the code directly when you change its value (as explained further in [1]).
Yeah, I don’t really see the point of it just glancing at the website. In fact, that little stack of activity notifications at the bottom seem quite obstructive and unintuitive, not sure if it’s part of the product but as a UI element it looks… off.
(It's not part of the free version since that works with your existing agent and you interact with it however you normally would).
The message stack in the bottom are the agent replies as it works. The idea is to move away from an array of messages as the primary interface with an agent and instead make the tools the primary interface. Whether that's actually helpful is an open question but it's certainly how I prefer to work now that I've been building with it.
It's an example of an algorithm that can be complex to read in code, but is more approachable when viewed as a diagram -- a perfect case for a tool like this.
Also thank you for your feedback and taking the time to view the landing page at least!
It's confusing enough that I don't know if this is just way above my knowledge level or whether I just don't understand what it even is.
It feels like you've developed something really useful but it's so internally obvious in your own head that it's missing some steps in the explanation for everyone else on how to come along with you and see the power of it.
I'm actually using Radix right now to brainstorm and update the example on the landing page to be more descriptive.
If it helps: the basic idea is that instead of an agent returning text or markdown we get it to return HTML. We then provide an app which makes working with that HTML actually useful (i.e. a sidebar to list the tools you've built, a way of commenting on the pages it generates).
The most trivial example I can think of is creating a simple colour picker. With Radix you can get an agent to add a colour picker that updates the code directly when you change its value (as explained further in [1]).
[1] https://radix-os.com/notes/hello
The message stack in the bottom are the agent replies as it works. The idea is to move away from an array of messages as the primary interface with an agent and instead make the tools the primary interface. Whether that's actually helpful is an open question but it's certainly how I prefer to work now that I've been building with it.
https://news.ycombinator.com/newsguidelines.html
https://news.ycombinator.com/showhn.html
Also thank you for your feedback and taking the time to view the landing page at least!