Humans spend most of their time completing tasks to satisfy their needs. However, when solving a task, the human brain executes a sequence of actions guided by a ‘task-set’, which is the configuration of processes that are initiated when a task is started and continues until it is finished or abandoned. Biological studies suggest that task-set processing mainly occurs in the prefrontal areas, although the mechanisms involved remain unclear. This article proposes a task-set model for plan retrieval, based on cognitive sciences. The objective of the proposed model is to improve the adaptability of computational systems in dynamic environments. The model was implemented and embedded in a task-set model to validate it through a case study, in this case, the match/no-match task, which consists of giving a specific answer according to the figures displayed on the screen.