Human Input Mode to allow choosing between Alway, Terminate or Never

[naseerfaheem]

in the human_input mode, can you please add the ability so the use can pick wether they want to have an input in langchain's thought or not? This will help steer agents' thinking in the right direction.

[joaomdmoura]

This is a great feature request actually, allows for course correct during the execution, adding as feature-accepted