Auto pausing and restarting a cluster on a schedule

Another thing that might interest you is the Atlas API. See

and

I presume that the above API and a cron job would do the trick if triggers do not accomplish what you want.