How much detail and how many levels should my work breakdown structure go to and how do I know if I'm done?
The general answer to this question is to go far enough down in levels and details to be sure your schedule estimates will have a good degree of accuracy. One rule of thumb is to take the WBS down to activity blocks of no more than 2-weeks' duration. But the level of detail, and what constitutes the right degree of accuracy, is ultimately a judgment call for the project manager. As you're breaking the work down, ask yourself questions like these
about going down to another level of detail in your WBS:
Would going deeper help me identify any major work that I have not already identified?
Would it help me identify any critical resources that I have not already identified?
Are project costs so critical that I might be missing significant costs at the current level of detail?
Is this turning into more of an exercise, and my planning time would be better spent on scheduling these tasks?
Remember that the schedule will generally go to a lower level of task detail than the WBS, decomposing chunks of work from the WBS into tasks of one to three weeks, to yield good task estimates and "hooks" for tracking and control.