Intelligent agent technology has evolved rapidly over the past few years along with the growing number of agent applications in various domains. However, very little work has been dedicated to define quality metrics for the design of an agent-based system. Previous efforts mostly focus on adopting classical metrics such as coupling and cohesion to measure quality of an agent design. We argue that the time has come to work towards a set of software quality metrics that are specific to the distinct characteristics of agent-based systems. In this paper, we propose a method to measure the reactivity of an agent design which provides indications of how the agent system responds to changes in the environment in a timely fashion. The proposed metric is part of the framework which facilitates the restructuring of an Belief-Desire-Intention agent program to improve its reactivity. Our framework was developed into a prototype tool which is integrated with Jason, a well-known agent-oriented programming platform.