I am curious how much of the functionality in this area is "write code that does scaffolding and prompt engineering" vs "finetuning and model-level improvements".
When I try to use it it asks me to sign in with GitHub, but after having signed in with GitHub and I try to use it again, it still asks me to sign in again.
The hamburger menu on your page shows that I am logged in, displaying my GitHub username and my email address.
The prompt I used was not long either.
First ux is pretty shitty. Taking time to create a prompt, getting redirected to login, login successfully, then lose the prompt. Surely you can preserve the prompt with a cookie or even in the url.
But the product is cool after that
Curious if you measure this and how much drop off there is between fill out a valid prompt and not logging in to see the results.