The 'core' code:
berlin.action <-- represents the basic game request and response cycle as expected by the game server berlin.servlet <-- main class and handling of raw HTTP request parameters, JSON etc. berlin.game <-- processes requests and responses into a higher level representation, for use by strategies
Places to look for customizing your AI:
berlin.setup.Setup <-- create all your strategy objects and the likes here. berlin.strategy <-- a nice place to put your strategies and strategy factories. There are some starter strategies in here already.