I'm writing a bot powered by wit.ai and want to figure out how to do it right. I followed the documentation, but there are some gaps.
I'm interested in particular case when in the first message I have only "intent" and in the second user message I have only "taste" (or two). Wit.ai api sometimes returns "stop" and doesn't jump to getCocktail execution.
How should I keep context of the dialogue to show that intent is still "cocktail"? Should I store it in context object?