Mine slopes under complex geological conditions and engineering disturbances commonly exhibit coupled characteristics involving deformation development, stability failure, and risk evolution, often accompanied by ecological degradation. From a resilience perspective, mine slopes are regarded as integrated systems that encompass structural safety, disturbance resistance, post-failure recovery capacity, and ecological restoration suitability. To address the insufficient linkage among prediction results, evaluation outcomes, and engineering decisions, this study develops a resilience-based analytical framework for comprehensive assessment and management of mine slope resilience. Within the proposed framework, a machine-learning-based slope stability prediction module is developed, and its workflow is illustrated in Fig. 1. The predicted stability states are integrated with deformation information, stability-related data, and resilience evaluation indicators to achieve multi-source information integration. In addition, traditional evaluation approaches are optimized using an improved radial movement optimization strategy, enhancing the integrity and interpretability of resilience assessment results and supporting risk classification, prioritization of control measures, and selection of mitigation strategies. Application to representative mine slope cases in the Yellow River Basin demonstrates that the proposed framework can effectively characterize slope resilience levels and ecological restoration suitability, thereby improving the applicability of assessment outcomes in practical engineering decision-making. The proposed approach provides a systematic and transferable methodology for mine slope resilience assessment and sustainable management.