Kyoto Fusioneering’s SCYLLA self‑cooled breeding blanket concept integrates SiCf/SiC structural composites with lithium-lead (LiPb, 15.7 at.% Li) to couple cooling and tritium breeding while targeting commercial‑grade reliability. We present a thermo‑mechanical optimisation workflow that begins with characterisation of anisotropic laminate properties and application of classical laminate theory to predict elastic coefficients validated against experimental data, enabling selection of a quasi‑isotropic layup for efficient structural simulation. Module stresses under hydrostatic loading are first approximated using beam‑bending formulations, then explored through parametric studies of rib reinforcement and wall thickness. A multi‑objective optimisation minimises ribs and wall thickness while maximising LiPb volume; stress reductions exceeding 95% are achieved without compromising tritium‑breeding space. The optimised geometry is generated in MATLAB numeric computing environment and validated in COMSOL multiphysics simulation software, showing strong agreement and significant peak‑stress mitigation. Transient survivability is assessed using RACLETTE, a code that simulates extreme heat transfer effects, to evaluate disruption‑driven thermal loads on the W-SiCf/SiC first wall interface for STEP‑relevant scenarios. Results highlight the critical role of high heat‑transfer coefficients at 700°C LiPb inlet temperature, the limited thermal benefit and added mechanical concern of W armour beyond 3 mm, and trade‑offs introduced by lowering LiPb temperature to 500 °C that relate to thermal margin vs irradiation‑induced swelling risk. This integrated composite‑mechanics, geometric‑optimisation, and transient‑response framework informs design choices that preserve breeding performance while delivering structurally robust, disruption‑tolerant SCYLLA blanket modules suitable for fusion power plant application.