This paper relates an experience in using a component model to design and construct agents. After discussing various rationales and architectural styles for decomposing an agent architecture, we describe a model of component for agents, named MALEVA. In this model, components encapsulate various units of agent behaviors (e.g., follow gradient, flee, reproduce). It provides an explicit notion of control flow between components (reified through specific control ports, connexions and components), for a fine grain control of activation and scheduling. Moreover, a notion of composite component allows complex behaviors to be constructed from simpler ones. Two examples, in the domain of multi-agent based simulation, are presented in this paper. They illustrate the ability of the model to facilitate both bottom-up and top-down approaches for agent design and construction and also to help at different types of potential reuse.