This manuscript explores the concept of goal adaptation in collaborative environments, specifically through the use of a simulation platform designed to train adaptive behaviors in response to unforeseen events. Drawing on the theory of graceful extensibility (TGE), the study aims to identify and formalize a pattern of goal adaptation in simulated contexts, which can be recognized and applied in real-world settings. Through the analysis of the simulation platform, two key events, a malfunctioning and a blackout, are shown to trigger shifts in the relevance of the system’s goals, prompting agents to modify their goal and plan accordingly. Albeit in a different context, the manuscript shows how a similar pattern can also be recognized in a manufacturing context. This research contributes to understanding how higher-order cognitive processes can be stimulated in simulative-based training, offering a foundation for future work in developing context-free tools for complex, and adaptive skills in various fields.