First, here's my primary use case why I purchased this:
1. Compare output from multiple models in one interface (occasionally... less common use)
2. Have a reliable multi-LLM powered Copilot that can store tons of my custom prompts with me in all interfaces where I work to speed up the work-flow and avoid paying for multiple subscriptions that I need to use only occasionally. (highly common use)

This is one of those AppSumo deals that I really, really root for and hope they don't drop the ball.

At the current stage it is usable but rough around the edges and isn't 100% reliable yet. Some connection errors here and there etc.

2 most important things for me that will make it or break it for me personally:
1. Reliable connections. As mentioned, I occasionally get 403 errors and network errors which interrupt my work-flow which is a big deal.
2. Improved custom prompt storage solution that would allow creating Categories and Subcategories that can be easily accessed both from Copilot and Playground interfaces. Current UI is fine for those who use a few prompts here and there to summarize, expand etc. I write lengthy, complex prompts for all sorts of use cases and at some point accessing and copy pasting them from docs is a major pain in buns. Current solution IMHO isn't going to work for those that work with large libraries with lengthy complex prompts like me.

I could nitpick about other things (I send my feedback to support email) but the rest IMHO are minor annoyances not real deal-breakers. These two are key things for me personally.

If before the refund clock is over they resolve these, I'll be singing praises from the rooftops as it will most certainly become my most used tool in my software toolbox. If not, I'll most probably refund and keep looking for an alternative.

If I may suggest, don't get distracted with introducing new features (yes, I'd like dark interface too but it's not critical), nail the fundamentals, please.

Rooting for you. Thanks.