Hey!

Hey @robertthomas2, thanks for the compliments. There are a few commands that you need to run to make it work. It's not complicated for a regular user, but it's meant to be a quick in-terminal LLM assistant for those who use CLIs the most like devs.

Thank you so much, @annedevj! I'm so much amazed that you found Hey useful. It can't compete with Warp's AI though. 😅 I use Hey in Warp. It's perfect. You can stay posted and get updates about Hey by following me on Twitter: https://twitter.com/lnxpylnxpy

Thank you, @star_boat! I really appreciate that. I developed the tool on my MBP. Then tested it on Windows and Linux environments. With a bit of modification, I made it compatible with all machines. It's POSIX-friendly.

Thank you, @muhammad_salman39! You can use Microsoft's Copilot for that purpose. It's designed for that purpose and gives you more accurate results.

Thank you, @elke_qin! That's lovely.

Hey @kyrylosilin, the handling that you mentioned is on LLM's and not Hey itself. You specify the LLM service URL and the specific model that you're trying to use, it'll then use that model to provide you the results in a pretty styled way. I use it in my daily work. It works perfectly fine!

@lnxpy thought that might be the case... kind of ironic given that this is an LLM powered CLI? :)

Great point, Chris! The LLM is just one part of the overall structure. When the LLM is in use, it takes a few seconds to respond, which isn’t ideal for handling built-in static commands that don’t need the model’s involvement. That’s why I structured 'Hey' to be versatile for both users and developers. Developers can install 'Hey' and leverage its internal APIs to integrate LLMs into their workflows. For example, I’m currently using 'Hey' (as a package) alongside an Arduino to control the lighting in my room—kind of like how we use 'Hey Siri' to control our Apple devices. (I accidentally deleted my previous reply—apologies for that.)