The ‘Open Source Community Sustainability’ project is about simulating open-source community interactions to encourage conditions leading to sustainable practice. I propose to do this by creating a framework with an agent-based model powered by Large Language Models (LLMs). This project has been worked on in OREL for the past 2 years, and a variety of models and tools have been created towards it, utilizing reinforcement learning, active inference and attraction-repulsion based approaches to build several agent-based models. Thus, this framework will represent open-source communities as a system composed of interconnected components (agents), mapping the intricacies of community engagement and collaboration as well as focusing on emergent properties and overall system behaviour. This project is important to the open-source ecosystem because as the demand for open-source software continues to grow, so do the challenges associated with community management, collaboration, and sustainability. Thus, providing a framework to understand how individuals collectively contribute to community growth and success can help derive insights into effective community management, collaboration strategies, and sustainability factors.