Summary. Due to the limited effective range of every physical sensor, as well as potential bandwidth limitations on the communication channels, there is a need for modeling and analysis tools when studying multi-agent robot systems that take these local interactions into account. In this paper, we review some of the work done on graphs and configuration spaces, and introduce the connectivity graph as a bridge between these two areas. We give sufficient conditions for a graph to be a connectivity graph in the sense that it can be realized as a formation in the configuration space. Moreover, the topological shape of a given connectivity graph is captured using triangulation techniques.