Abstract: In the current context of accelerating brandization of agricultural products, the supply chain of branded agricultural products is trapped in dilemmas such as information opacity, rampant fake reviews, and lack of brand credibility, severely hindering the healthy development of agricultural product brands and weakening their market competitiveness.To effectively address these challenges, this article proposes a credit management solution based on blockchain and virtual cooperatives, aiming to build a transparent and trustworthy credit management system for agricultural product brands, which will help enhance the market credibility and competitiveness of agricultural product brands. An in-depth analysis of the social relationship network in agricultural product production shows that the production of branded agricultural products follows a "government regulation - brand operator-led - multi-entity collaboration" model, involving multiple stakeholders such as farmers and cooperatives. Based on this network, a virtual cooperative trust management model under blockchain is constructed. A density peak clustering algorithm is introduced to filter out false information, and a dual-layer credit quantification system is built using topological potential theory. A credit-weighted mixed consensus algorithm is designed, and a blockchain structure that coexists with traceability information and credit management is planned. In the filtering of false information, the introduced Density Peak Clustering (DPC) algorithm plays a key role. This algorithm determines the cluster centers by calculating the local density of data points and their distance to high-density points. In the context of agricultural product reviews, it can accurately locate the cluster centers with the help of strategy graphs, effectively identifying malicious reviews even with few parameters. Malicious reviews, characterized by extreme content and repetitive expressions, are aggregated and removed by the algorithm, preserving the true preferences of the real group and ensuring the authenticity and reliability of the review data. Credit calculation is a core link, and the dual-layer credit quantification system constructed using topological potential theory has achieved remarkable results. Within the community, the influence of nodes is measured based on factors such as interaction frequency and quality with other nodes, thereby quantifying individual credit, with higher influence nodes having higher credit. Between communities, the group credit radiation effect is evaluated through the topological potential field, visually demonstrating the degree of mutual influence of credit between different communities, achieving coordinated measurement of micro individual and macro group credit. The credit-weighted mixed consensus algorithm designed in conjunction with the characteristics of virtual cooperatives comprehensively considers factors such as node credit levels and computing power to allocate weights, allowing nodes with high credit and strong computing capabilities to have greater influence in the consensus process, thereby improving consensus efficiency and security. After reaching a consensus, the credit data is stored in a blockchain credit block, ensuring immutability and traceability. According to the needs for agricultural product quality and safety traceability and credit management, a single tree multi-type transaction storage block structure that accommodates both traceability information and credit management is designed in conjunction with the storage characteristics of Hyperledger Fabric, meeting actual business requirements. Experiment results show that this credit model is sensitive to false information and the credit of malicious nodes improves slowly. When the proportion of malicious nodes reaches 50%, the recognition rate still remains above 56%. In a gigabit local area network environment, the CWHC consensus algorithm maintains stable consensus delay in a 4-consensus-node network with 1 abnormal node. When there are 60 transaction records, the communication volume is reduced by nearly 10,000 kb compared to the practical Byzantine fault-tolerant mechanism. This credit management solution based on blockchain and virtual cooperatives effectively addresses the credit management challenges in the brand agricultural product supply chain and provides a feasible new path for agricultural product brand credit management.