spade 2.3.2

Smart Python multi-Agent Development Environment

Welcome to SPADE

SPADE (Smart Python multi-Agent Development Environment) is a Multiagent and Organizations Platform based on the XMPP/Jabber technology and written in the Python programming language. This technology offers by itself many features and facilities that ease the construction of MAS, such as an existing communication channel, the concepts of users (agents) and servers (platforms) and an extensible communication protocol based on XML, just like FIPA-ACL. Many other agent platforms exist, but SPADE is the first to base its roots on the XMPP technology.

The SPADE Agent Platform does not require (but strongly recommends) the operation of agents made with the SPADE Agent Library. The platform itself uses the library to empower its internals, but aside from that, you can develop your own agents in the programming language of your choice and use them with SPADE. The only requirement those agents must fulfill is to be able to communicate through the XMPP protocol. The FIPA-ACL messages will be embedded in XMPP messages. Be warned, however, that some features of the whole SPADE experience may not be available if you do not use the SPADE Agent Library to build your agents.

The SPADE Agent Library

The SPADE Agent Library is a module for the Python programming language for building SPADE agents. It is a collection of classes, functions and tools for creating new SPADE agents that can work with the SPADE Agent Platform. Using it is the best way to start developing your own SPADE agents. In the future, we would like to offer bindings of the SPADE Agent Library for more programming languages, like Java or C#, but for now only Python is supported.

If you have worked with Python before, you will find the SPADE Agent Library easy to understand and use, with classes ready to use and expand, and many functionalities already built in the library. If you have worked with other high-level behavioral agent platforms (like Jade or MadKit) you will find some similarities in the way the library works, although the agent model present in SPADE is a bit different.