Capability modeling starts at the end of the architectural design when system sub-actors have been specified in terms of their own goals and the dependencies with other actors. In order to define, choose and execute a plan for achieving its own goals, each system’s sub-actor has to be provided with specific ‘‘individual’’ capabilities. Additional ‘‘social’’ capabilities should be also provided for managing dependencies with other actors. Goals and plans become integral part of the capabilities. In detailed design, each agent’s capability is further specified and then coded during the implementation phase.