Assessing Hypothesis Testing (HT) in undergraduate inferential statistics presents a persistent challenge. The concept itself is not simple and can lead to misinterpretations by students, such as misinterpreting the significance level and failing to recognize its influence on the test; here, two approaches may clash—those of Fisher and Neyman-Pearson—and sometimes it is the instructors themselves who do not recognize their difference. Batanero et al. [3, 4] and Vallecillos [12] have addressed this and many other types of difficulties that arise in the teaching and learning of the concept of hypothesis testing (HT), demonstrating in their work the need to improve methods and assessment in courses where it is covered. This paper proposes that assessment consider the complexities and nuances of the concept, from a didactic approach centered on the design of test items aligned with the dimensions of cognitive competencies: interpretative, argumentative, and propositional. All of this is supported by the R package exams, which stands out as a robust, open-source tool for the automatic and customized generation of assessments. The main goal was to show how the R exams package contributes to better assessment design with customized, reliable, and structured questions, aiming to foster a deeper and more meaningful understanding for students of the concept of hypothesis testing.