[Research] Programing and with starlogo

Digging into the world of modeling and simulation, I tried the starlogo tutorial: programing ants to search for woodchips and place them into piles: This is cool but I am wondering if I can program more complex behavior. The other issue is that what I want to simulate is not a decentralized behavior. 3 agents collaborating might have a leader. Furthermore, I am in a context where I want to investigate the behavior or 3,4,5 agents and not 1000...