Made a script to enable mute/unmute of my mic via a custom button on my keyboard.
Just an observation about coding w/ #ai I specify what I want, get result it is sure will work, find errors and/or it's not quite what I want, post results and state issues to be fixed -- rinse, repeat.
I noticed during this endless iterative cycle that #AI would re-introduce bugs or misunderstandings we fixed previously. I assume this is due to the conversation aging and token limits. Finally, I got things to work by telling it to remember a numbered list of requirements.