Networks of multi-agent systems are considered to be heterogeneous if they include systems with different sets of APIs, running on different virtual machines. Developing an agent that can operate in this kind of a setting is a difficult task, because the process requires regeneration of the agent’s executable code, as well as modifications in the way it communicates with the environment. With the main goal of providing an effective solution to the heterogeneous agent mobility problem, a novel agent-oriented programming language, named ALAS, is proposed. The new language also provides a set of programming constructs that effectively hide the complexity of the overall agent development process. The design of the ALAS platform and an experiment presented in this paper will show that an agent written in ALAS is able to work in truly heterogeneous networks of multi-agent systems. [Acknowledgements. This work is partially supported by Ministry of Education and Science of the Republic of Serbia, through project no. OI174023: ”Intelligent techniques and their integration into wide-spectrum decision support”.]