The multi-agent framework is split up into different parts. Each component is provided with separate documentation. In this section, we give an overview of all framework elements. The whole framework is published at Maven Central.
The source codes are written in Java (latest release version) with a defined syntax structure; all code documentation is created with Doxygen. We are also using unit testing and the Git-Flow developing model.
|AgentSpeak(L++)||The component is before the first release, we are in the testing phase and increase test-cases to find any problems, all necessary elements exist and are mainly tested||This component defines the main agent structure. It defines the whole syntax and the base agent structure. It also supports an agent generator and configuration so that you can create a set of agents by an ASL-file. Based on the interface structure, you can redefine all elements if you like.|
|REST||The component is currently under development, the main functionality is implemented, but the system is not tested completely and some features can be missing||The REST-API (ReSTful) is a system to define communication over HTTP for distributed systems. We are supporting such a REST-API to control a set of agents and agent generators. So the whole multi-agent simulation can be controlled via a web interface based on a servlet container on a Java web server e.g. GlassFish, Jetty or Tomcat|
|Scratchpad||The componentis started up, so at the moment the basic infrastructure will be build||The idea of a scratchpad depends on our workshop, hands-on tutorials and on our wizard. We are building a high-scalibility cloud-based application of a multi-agent system with current state-of-the-art technologies like Spring-Boot, Netflix Hysterix and React|