Hello :)
=======

FLUXY is a lightweight Python Multi Agent Framework. It abstracts from the concrete message transport layer, thus
more or less any message oriented middleware can easily be used with it. Just extend the "abstract"
MTL/MTLAgent classes. For the moment it comes with a binding to ZeroMQ.

Being inspired by MaDKit FLUXY is an Organization Centric MAS solution (OCMAS).
Similar to MaDKit it implements the Agent/Group/Role (AGR) model.
Any number of agents can be members of any number of groups where they can have an arbitrary number of roles within.

FLUXY also enables a consequently asynchronous programming style (Closures as callbacks and so on).

Agents in FLUXY have Identities referring to the roles the agents play in groups. Associated with each Identity of
an agent is a Behaviour which consists of ActionPatterns and ReactionPatterns.

There are InformMessages (one-way messages) and Request-/RequestReplyMessages (request-respond messaging) in FLUXY.
You can derive application specific message classes from these base classes and register them at the Message Parser.
Currently FIPA-ACL message classes are not implemented yet. However, this will be done as soon as I find some time ;)

There is a minimal internal DSL for the messaging use cases that should help to keep the code readable.

FLUXY is still in ALPHA-stage.


Requirements:
============

- Python 2.7
(- ZeroMQ / pyzmq)


Usage:
======

Have a look at examples/prime_numbers.py ...

To start the platform execute the start_platform.py script.

Some configuration can be done in the Config.py file.


Questions:
=========

Simply send me an email: martin.beckmann (at) evolving-algorithms.org (Martin Beckmann)


[ More documentation and extensions coming soon ;) ]