Multiagent systems (MASs) courses are nowadays present in most master's degrees in Artificial Intelligence. In such courses, students have to cope with the gap between some high‐level theoretical concepts and techniques that allow agents to reach agreements in order successfully cooperate (the so‐called Agreement Technologies) and their corresponding implementations to be used in real programs. If the learning outcomes of such courses include the development of actual (although reasonably simple) MASs, the incorporation of some strategy to make students enhance their initial programming skills becomes crucial for their success. In this context, this article presents both a tool called SimFleet and a teaching methodology that have been designed together to guide students throughout a typical MASs course and, specifically, to enhance their programming skills in order learn how to develop actual MASs.