This proposal investigates how settlement can coexist with landslide-prone landscapes by working with the processes that shape them rather than attempting to resist these processes permanently. Instead of positioning architecture as a direct response to instability, the research develops an adaptive settlement framework for environments where water, ground, and occupation remain in continuous transformation. Landslides are understood as part of a dynamic territorial condition, raising the question of how hydrological intelligence can guide where, under what conditions, and with what degree of permanence settlement can occur.  Escazú, Costa Rica, serves as the testbed. Its steep terrain, intense rainfall, and expanding urbanisation create conditions in which altered drainage can increase soil saturation and landslide susceptibility. The territory is read through ecotonal gradients, distinguishing areas that require protection, progressive stabilisation, or occupation. Computational hydrological simulations trace water movement through the watershed and inform a branching network of channels, terraces and collection landscapes that follows the existing topography rather than overriding it. Material research extends this hydrological logic into the construction system. Rice-husk ash, lime and locally available aggregates are developed into materials with differentiated water and mechanical behaviours: hydrophobic mixtures form drainage channels that convey water, while hydrophilic mixtures define collection nodes that retain and gradually absorb it. This hydrological gradient is combined with bioremediation and planted terraces, which progressively stabilise vulnerable ground, reduce erosion, and restore degraded areas. Agricultural terraces extend this process into productive landscapes, linking ecological recovery with local cultivation and supporting an agro-urban settlement strategy.  Architecture responds to these changing landscape conditions through varying degrees of permanence. Adaptable bamboo structures occupy areas where environmental change remains possible, while permanent buildings develop along more stable contours and terraces. Through the integration of computational hydrology, bioremediation, material experimentation, and architectural design, the project proposes a transferable framework for determining how settlement can adapt to the changing conditions of a dynamic landscape. Resilience is therefore understood not as permanently fixing unstable ground, but as allowing settlement to evolve according to the changing capacity of the landscape to support it.