Initiated by significant carbon emissions of accelerating urbanization, there has been increasing interest in large-scale additive manufacturing (LSAM) and functionally graded design (FGD) in architecture. These innovations prove the potential to speed up, simplify and optimize design pipelines as well as manufacturing processes to contribute to a sustainable future. Yet, conventional computer-aided design software frameworks lack facilitating and explicit modeling environments to utilize LSAM and FGD in the industry. Furthermore, these domains get commonly separated because of their difference in design scale. This dissertation outlines a novel design framework combining LSAM and FGD. It introduces functionally graded, differential-grown, and discretized 3D toolpaths of compression-only shells for extrusion-based LSAM. The framework builds on this research's outlined algorithm for graded differential growth (GDG), introducing graded point relaxation. The dissertation also contains an overview of recent developments and design frameworks for LSAM and FGD in an architectural context.