get started with salt
authenticate a paid inference provider
choose a model for inference
clear memory in the current channel, thread, or direct message
set the current channel for speech
choose a voice for speech