Porous architectures offer a highly adaptable geometric platform for regulating coupled physical processes, including heat transfer, fluid transport, and reactive flows. In complex energy and thermal systems, these fields interact strongly and are highly sensitive to structural topology. Recent studies have demonstrated that representative porous architectures, such as TPMS-based media, enable efficient modelling of coupled heat and mass transfer through geometry-controlled transport pathways [1], while porous media combustion systems exploit topology-induced internal heat recirculation to stabilise lean reactive flames [2]. In parallel, optimisation-driven mechanical cloaking has shown that physical field signatures can be suppressed through geometry alone, without reliance on material anisotropy [3]. The central objective of this work is to develop a systematic optimisation framework for porous structures that enables controlled manipulation of coupled multi-physics fields, with particular interest in achieving cloaking-like behaviour through geometry-driven design rather than material anisotropy or external field control. The study focuses on the generation and optimisation of three-dimensional porous architectures with spatially varying porosity, connectivity, and curvature. A voxel-based geometric framework is employed to explore a broad topological design space, allowing continuous modulation of internal transport pathways and field interactions. Multi-physics numerical simulations are used to evaluate the coupled responses of thermal, flow, and reactive (including combustion) fields within candidate structures, forming the basis for identifying structure–field relationships and guiding topology optimisation. The results demonstrate that appropriately designed porous topologies can redistribute coupled fields around prescribed regions, suppress field penetration, and induce cloaking-like effects without reliance on anisotropic materials or external field control. Overall, this work establishes geometry-driven design principles and provides a general, extensible methodology for porous-structure optimisation in multi-physics environments, with broad relevance to thermal management, flow control, and reactive energy systems.