Among all 3D structures, saddle-shaped geometries and minimal surfaces have intrigued scientists for centuries, because they can spontaneously self-assemble driven by the minimization of thermodynamic quantities such as the interfacial energy. Echinoderms, like sea urchins and sea stars, build a calcite skeleton whose microstructure is an outstanding biogenic example of saddle-shape geometry. This 3D structure, called stereom, is a porous meshwork made of calcite, whose peculiar curvature signature is conserved across different species and body parts. Several studies have addressed the morphogenesis of the stereom showing that it forms via biomineralization through a rich dynamics of branching and bridging episodes. Yet, a global, mechanistic comprehension of what controls preferential mineral deposition is still lacking. Recently, we have found that the stereom may inherit its specific geometry from a precursor made of actin-fibers that self-organize under tension close to the growing boundary. However, precursors are confined in a pore-size layer closed to the growing boundary and cannot explain the coherence of the overall network. Our hypothesis is that the stereom local geometry is not only an emergent property but also a driving cue for skeletonizing cells and that this interaction is mediated by the cytoskeleton. The internship aims to build a model, coupling the stereom geometry and the self-organization of actin fibers using a continuous approach.