The efficient transport of luminal contents in the small intestine is governed by a closed-loop control system linking sensory feedback (the sensor block), neural regulation (the controller block), muscular actuation (the actuator block) and fluid-structure interaction (the process block). Existing computational studies have typically addressed isolated blocks of this loop, such as fluid dynamics under prescribed motor patterns, offering limited system-level insights. Here, we present CREST (Closed-loop REgulation of Small intestinal Transport), an integrated mechano-physiological control framework for simulating small intestinal transport. CREST couples a mechanical module, which captures wall-deformation-driven fluid flow under muscle contractile force, with a physiological module, which senses strain, compares it to a reference strain and activates smooth muscle contraction via the enteric nervous system. Through this closed-loop interaction, CREST reproduced clustered peristaltic waves with amplitudes and velocities consistent with experimental observations, and revealed interesting transport phenomena that align with ex vivo data. By closing the feedback loop, CREST provides a powerful computational framework for system-level exploration of intestinal functions, laying the foundation for the digital intestine twin and bioinspired control strategies.