<p>An increasing number of slope failure incidents and engineering case studies underscore the limitations of relying solely on a safety factor (SF) to identify unstable rock masses and accurately categorize such formations. To address these limitations, this research introduces three indicators: the traditional safety factor (SF), cohesive safety factor (CSF), and cohesion-to-shear ratio (CSR). A methodology based on these three indicators is proposed for mechanically delineating sliding-type rock masses, allowing a more precise classification of hazardous formations. Simulated freeze–thaw experiments yield results demonstrate that rock masses transition into unstable states when their CSF falls below 1. Similarly, latent damage and deterioration trends within rock masses become evident when the CSR falls below the ratio of the rock mass's long-term strength to ultimate strength. Practical engineering cases, such as the Yaoziyan unstable rock mass in Chongqing, China, further validate the feasibility of the proposed identification method. This study presents a novel mechanical delineation methodology tailored to unstable sliding rock, overcoming the limitations of relying solely on a single SF in practical engineering contexts. This quantitative mechanical identification method improves the accuracy and scientific rigor of slope unstable rock mass identification, providing a more nuanced classification for hazardous formations.</p>