In the realm of Multi-Agent Reinforcement Learning (MARL), the challenge of ensuring effective coordination indifferent multi-agent teams remains a significant hurdle. Existing methods often fall short in generalizing learned policies tonovel team compositions, sizes, and capabilities. Addressing this gap, our study focuses on systems with variable and obscureattribute compositions, harnessing a context-based meta-reinforcement learning framework. The approach is twofold: contextinference and context-based decision-making. Agents interpret historical data to identify the system’s attribute composition,guiding their collective efforts. The accuracy of these inferences is crucial, prompting us to integrate contrastive learning torefine the context inference network via unsupervised training. In the process of decision-making, agents integrate the inferredcontext with the observation features to select the optimal strategy. Our empirical results underscore the method’s efficacy inbolstering decision-making efficiency and precision amidst attribute diversity, marking a significant stride in adaptive teamingfor robust multi-robot deployments in dynamic real-world scenarios.