In this paper we propose an extension to the traditional model of bounded rationality and
incorporate the extended model into a theory of organizational behavior. We argue that
organizations are collections of tasks and intelligent agents engaged in performing those tasks,
both situated within an organizational setting. Organizational behavior is an emergent property
of such collections and is constrained by the agent, the task, and the situation. We propose that a
unified theory of organizational behavior is possible, but only if agents, tasks, and situations are
specified at a sufficient level of detail, and only if that specification embodies both the agents'
mental models of the task and social-situation and the task and social-situation. Inattention to
relevant details of the agent, task, or situation (and their interactions) may produce misleading
results. We describe a candidate theory, ACTS theory, that integrates Agents who are
Cognitively-restricted, Task-oriented, and Socially-situated in an interlinked set of
representational systems. We suggest that the complexity of the theory warrants its realization
and testing in a computational form, and that there exist candidate computational theories of
cognitive agents and organizational situations. We illustrate the importance of attending to task
and the benefits of multi-point focus on agent cognition, task, and situation using two
computational models that take the ACTS perspective, yet are currently only partially sufficient
as models that embody a full ACTS theory.