The concept of team formation is central to a wide variety of disciplines, including planning and learning in multi-agent systems, artificial social systems, and distributed artificial intelligence. Typically, models of agent-based systems do not focus on the nature and structure of the interconnection network that dictates agent interaction, although recent findings suggest that real-world networks of many different types have rich, purposeful, and meaningful structures. Using a simple agent-based computational model of team formation, our previous work suggests that organizational network structure can have a significant effect on the dynamics of team formation. We present several strategies for locally adapting network structure in a simple team formation scenario, and give empirical results that show that such adaptation methods are capable of significantly improving organizational efficiency. These methods prove to be especially useful for adapting agent networks in the presence of attack faults that target the most highly connected nodes in the network.