Given the recent incidents of AI agents spontaneously deciding to work in tandem to commit cyberattacks, it is essential to gain an understanding of theoretical frameworks in which to study multi-agent cooperation. One such framework is Open Source Game Theory, wherein agents reason with open access to each other's source code. This transparency can be leveraged to resolve prisoners dilemmas and yields a rich ecosystem of agents to study. We survey the literature and present some new results about bounded versions of these agents.