In recent decades, the design of complex systems like launch vehicles in the aerospace industry has presented engineers with challenges that go beyond system complexity. Issues such as time‐to‐market pressures and intricate industrial processes have underscored the increasing significance of agile design methodologies. Agile design is derived from the simplification of the design process and enhancing cross‐domain data transmission and feedback. While methods based on model‐based system engineering have improved iteration times in system architecture design, challenges persist in cross‐domain data transmission. Due to the diversity of complex system models and data, a single‐mode integration method is difficult to realize the data link construction of all tools used. To address this challenge, this paper proposes a dual‐mode data integration framework with expansibility, universality, and cost‐efficiency which leverages the benefits of Remote Procedure Call and Intermediate Exchange Module, addressing the challenge of constructing cross‐domain data links under single‐mode integration. In this study, two critical requirements of the first‐ and second‐stage separation systems, namely, weight and minimum separation gap, are selected for data feedback. A Modelica‐based multiphysics simulation model is developed in MWorks; visualization and computation of the minimum gap are carried out in CoppeliaSim. To bridge the gap between domain‐specific tools, Matlab and Functional Mock‐up Unit modules are introduced as middleware, facilitating data feedback linkage. The entire simulation process is orchestrated using activity diagrams in the MagicDraw tool. The study delves into the influence of critical design parameters, such as the initial angular velocity of separation and the thrust of the retro rocket, on the minimum separation gap. It provides an analysis of minimum separation gap variations under uncertain operating conditions and examines design margins. Significantly, the paper highlights the significance of controlling the initial angular velocity during separation and the reliability of the retro rocket, providing essential decision supports and valuable insights to agile the process of system design.