Abstract: To address the integrated optimization of project scheduling and personnel staffing problem under resource constraints and uncertainty, a three-stage robust optimization framework integrating adaptive large neighborhood search (ALNS) algorithm, starting time criticality (STC) and the VIKOR method is proposed. The framework aims to achieve a systematic trade-off among project duration, personnel cost and robustness. The process is divided into three stages: First, an ALNS algorithm is employed to jointly optimize activity execution modes, activities priorities and personnel allocation, generating a set of candidate solutions that balance project duration and cost. Second, the framework innovatively extends the STC strategy from its traditional application in duration uncertainty to scenarios of work content uncertainty, enhancing solution robustness through the proactive allocation of time buffers and quantifying the associated robustness cost. Finally, the VIKOR method is applied to comprehensively rank the robust solutions based on criteria of project duration, personnel cost and robustness cost, thereby identifying the optimal compromise solution. Numerical simulations and sensitive analyses demonstrate that the proposed framework effectively mitigates the impact of disruptions, exhibiting significant advantages in both robustness and overall performance compared to benchmark methods. It confirms that the three-stage robust optimization framework provides an efficient, reliable and practical systematic approach for decision-making in complex project environments, capable of identifying solutions that achieve an optimal trade-off among conflicting objectives.