The selection of chemical reactions is directly related to the quality of synthesis pathways, a reasonable reaction evaluation index plays a crucial role in the design and planning of synthesis pathways. Since the construction of traditional reaction evaluation indicators mostly rely on the structure of molecules rather than the reactions themselves, considering the impact of reaction agents poses a challenge for traditional evaluation indicators. In this study, we first propose a chemical reaction graph descriptor that includes the mapping relationship of atoms to effectively extract reaction features. Then, through pre-training using graph contrastive learning and fine-tuning through supervised learning, we establish a model for generating the probability of reaction superiority (RSscore). Finally, to validate the effectiveness of the current evaluation index, RSscore is applied in two applications: reaction evaluation and synthesis routes analysis, which proves that the RSscore provides an important agents-considered evaluation criterion for Computer-Aided Synthesis Planning (CASP).