Social network (SN) holds significant sway over the consciousness and behavior of involved individuals, serving as an evolutionary medium for opinion dynamics models. Consensus is a fundamental concept in group decision-making (GDM). How to effectively explore the opinion evolution during the consensus-reaching process through SN is of paramount significance for decision-making science. Therefore, a dual-mechanism containing trust exploration and leadership incubation is developed for modeling opinion dynamics, creating a favorable condition for achieving consensus. First, a novel mechanism for analyzing the completeness of SN is proposed, encompassing a trust propagation process that considers trust discount and stability, as well as a trust aggregation method grounded by quantum theory. Second, a trust screening rule is discussed to retain the valid indirect trust relationships (TRs), and then a leadership incubation mechanism is developed to promote the effective achievement of consensus opinions in group decision-making. Finally, a numerical study is presented to elucidate the superiority and rationality of the proposed methods, and some simulation experiments and comparative analyses demonstrating the effectiveness and advantages of which are covered.