A Sustainably Evolvable Immune System Simulation Platform: An Engineering Framework and Implementation for a QSP / Systems Immunology Credibility Loop

Protocol preprint (v1.0; no measured numbers). We present an immune-system simulation platform for sustainably evolvable (long-horizon, perturbation-rich) immune dynamics whose contribution is the engineering of QSP / Systems Immunology credibility elements into an executable loop: Context-of-Use (COU) statements with risk grading; explicit measurement models; strict isolation of calibration, uncertainty quantification (UQ) and extrapolative validation; joint coverage-and-sharpness constraints on UQ (WIS/CRPS); graded extrapolation (L1 weak / L2 strong); and Verification-lite discretization-sensitivity checks. A unified kernel (ABM + global ODE + 2D reaction-diffusion) and scenario packages cover three murine baseline scenarios (respiratory virus, bacterial infection, tumor immunology), with a unified replay protocol and audit trail for click-through review. Results are given as protocol and figure placeholders; empirical tables will be released as a New version whether positive or negative. Files: English and Chinese manuscripts (MD + PDF), reference helper immunesim_ref.py, LICENSE, CITATION.cff.