In the increasingly competitive manufacturing industry, optimizing production decision making and quality control is crucial for the strategic development of companies. To maximize cost effectiveness and enhance market competitiveness, scientific decision-making and effective quality inspection are particularly important. Among the various types of decision models for production processes, extensive research has been conducted in different fields to address diverse decision problems for production processes, resulting in the establishment of multiple models that aid in the analysis of factors that influence processes at various stages. In this paper, we propose a production decision optimization method based on a multi-agent mixed-integer programming model, which integrates multistage decision analysis and quality inspection. By incorporating Monte Carlo simulation, we can simulate the fluctuations in defect rates during actual production processes and optimize decision-making under multiple confidence levels. This model effectively balances production costs and product quality, achieving maximum cost-effectiveness through the optimization of decision pathways during the production stages. Experimental results show that our model can provide robust and efficient decision support in dynamic manufacturing environments.