Purpose As artificial intelligence absorbs technical tasks, competence is being redefined around what remains distinctly human: patient safety, quality of care, interprofessional collaboration, bioethics, and self-care. Yet undergraduate assessment still compresses these into a single episodic score. We are building a clinical-case instrument that profiles five transversal domains separately, giving learners growth-oriented feedback instead of one metric. Methods This exploratory work in progress pilots three clinical cases with five multiple-choice items each (15 items, 20 minutes). Four features shape the design: each case opens in the patient's own voice before any clinical framing; every character carries a full name and a socioeconomically diverse profile; gender roles are distributed equitably; and each item activates two or three domains at once. Content aligns with Mexican health regulation and the WHO Framework for Interprofessional Education. Participants will be third-year or higher medical students at a Mexican university (convenience sample; minimum 30, target 60-100), recruited through open institutional announcements. Analysis includes descriptives by domain, repeated-measures ANOVA or Friedman, inter-domain correlations, and exploratory Cronbach's alpha by subscale. Results (work in progress) Instrument construction and the specifications table are complete; every item is mapped to its target domains and to national safety norms. Ethics review is in process; administration begins September 2026. The talk will present the design rationale, the specifications table, and early data where available, and invite discussion on scoring items that load on multiple domains. Conclusion Integrated, patient-voiced cases may make transversal competence visible as a profile of strengths and gaps rather than a single pass/fail judgment, the kind of evidence that supports mentorship, targeted feedback, and second chances.