Abstract: The reuse of mine underground space is an important pathway for safeguarding national deep-underground energy security and achieving the carbon peaking and carbon neutrality goals. However, unclear hazard-forming mechanisms and insufficient safety assurance remain key constraints on the reuse of mine underground space. Focusing on the theoretical and technical bottlenecks underlying the safe and integrated reuse of mine underground space, this study systematically reviews the current research status of disaster types, safety retrofitting, intelligent monitoring, hazard early warning, and safety assurance platforms for mine underground space reuse. The main findings are as follows. According to safety function, hazard-forming mechanism, and disaster manifestation, disasters associated with mine underground space reuse are classified into three categories: large-scale roof instability, engineering-disturbance-induced disasters, and sealing-failure-induced disasters. The strength-weakening and energy-imbalance evolution characteristics of large-scale roof instability, the cyclic-loading and fatigue-deterioration response characteristics of engineering-disturbance-induced disasters, and the fracture-coalescence and permeability-jump failure characteristics of sealing-failure-induced disasters are further examined. A safety retrofitting technology system covering roadway lining reinforcement and fractured rock mass grouting restoration is established. A lining design method based on the response characteristics of bedding structures and a refined grouting process based on a fracture zoning model are proposed. Intelligent monitoring technology for the safety of reuse spaces under full-scenario conditions is developed. A coupling relationship model between induced signal characteristics and disaster magnitude is established, and a multi-source heterogeneous monitoring network centered on microseismic monitoring and distributed optical fiber sensing is formed. A preliminary framework for a digital-intelligent safety assurance platform for mine underground space reuse is constructed. A four-dimensional spatiotemporal data cube integrating multi-source heterogeneous geological and monitoring data is established, and a multimodal information fusion and intelligent disaster risk prediction method based on random network graphs is proposed. These results provide theoretical and technical support for the safe implementation of the national deep-underground strategy.