This dissertation explores the topic of designing space-making robots, which are robotic envelopes and volumes (robotic walls, ceilings, floors, partitions, building skins, vehicle interiors, etc.) that define and reconfigure architectural spaces. Space-making robots are essential parts of smart built environment (SBE), architectural robotics, and smart city. They create, define, and reconfigure the architectural spaces we live in. Informed by HAI (Human Agent Interaction) and HRI (Human Robot Interaction) literature, the author argues that people have the tendency to perceive adaptive and interactive space-making robots as agents, so that they can be designed as our companions, partners, friends, etc. Consequently, the human-robot interactions of space-making robots can be designed as human-agent interactions. This design approach is proposed in this dissertation as the “HAI-based design paradigm for space-making robots,” which is validated through both theoretical lens and empirical studies. Based on this design paradigm, the author then proposes a patterned-based, design framework for collaborative environments as an exemplary application of this design paradigm, which is then validated qualitatively through a design exemplar of a partner-like, collaborative space-making robot. Since “space-making robots” are becoming more and more prevalent in our lives especially in confined spaces, the proposed design paradigm and framework can be widely applied to the design of SBE, architectural robotics, and smart cities.