The paper formalizes the Planning–Modeling–Prediction (PMP) methodology as a basis for clinical decision support systems used in preoperative planning in traumatology and orthopedics. Two anatomical regions are considered: the spine--pelvic complex and the first ray of the foot, for which computational, software, and methodological results had been obtained previously. Comparison of these implementations made it possible to identify the common elements of the digital workflow: clinical and radiological input data, a set of admissible surgical options, a biomechanical model, computed indicators, criteria of geometric and mechanical admissibility, predictive interpretation, and a rule for comparing treatment options. A formal description of the CDSS architecture is proposed, separating the general sequence of transformations from anatomy-specific content. The implementation for the spine–pelvic complex is shown to be more complete in terms of modules and accumulated data, whereas the implementation for the first ray of the foot follows the same architectural logic with changes in the radiographic basis, intervention parameters, loading scenarios, and interpretation of the result. The proposed scheme can be used to transfer the PMP methodology to other problems in musculoskeletal surgery, provided that features, constraints, and admissibility criteria are specified separately for each new anatomical region.