We live in a world increasingly digital where intelligent and communicating objects evolve and interact. These objects have their own missions and goals. However, in a more and more complex environment, mechanisms of communication and information exchange protocols need to be more efficient, fast and smart. The collaborative approach seems to be the most suitable to facilitate interaction between intelligent entities while preserving their resources by sharing relevant information that will enable them to reach more quickly their goals. For example, collaborative behaviours and informations exchanges could improve the movement of vehicles in an urban center and avoid traffic jams and more globally for all kind of autonomous agent. This article highlights what is a collaboration in the context of communicating autonomous entities. Then we present our architecture that is called HACCA (Hybrid Architecture for Collaborative Communicating Agent). We show the different levels implied in HACCA and we put the stress on the multigraph structure we have created to build and share contextual knowledge. The theory is applied to an example on communication between communicating vehicles.