In this paper, we propose an information exchange framework to incorporate smart building demands into a microgrid operation. We especially focus on scheduling smart building demand along with other possibly highly uncertain microgrid generation resources a day ahead or shorter ahead of operation. Since every building demand has different energy consumption profiles and dynamics, it is inefficient for the grid operator to directly control a potentially large number of buildings. On the other hand, if the grid operator does not have sufficient information on the buildings and their resource potentials, smart buildings cannot be deployed to the full extent. Our proposed framework attempts to align these conflicting objectives in a reasonable way, taking into consideration the realistic limitations of communication time, various physical characteristics and dynamics of the buildings, and the building managers' economic objectives. This framework is in line with the concept of dynamic monitoring and decision systems (DYMONDS), where the economic objectives and physical dynamics of any flexible generation and loads on the grid are aggregated by load aggregators and incorporated in the optimal system operation.