This study developed a designer-oriented smart home device computer-aided design tool that can help architectural designers to design smart home. This design tool is intended to incorporate smart home devices into overall spatial planning from the beginning of the planning process. Cloud computing  facilitates opportunities for online collaboration and coordination between individuals, between individuals and objects, and between objects, and will help speed the evolution smart home from centrally-controlled intelligent buildings to "smart living space" emphasizing distributed computation ability. This study employs (1) the theory of intelligent agents and (2) scenario-oriented design as an integrated framework for computer-aided design of smart home device. Smart home device are redefined based on agent theory. Agents' behavioral models include the five types of passive, reactive, active, interactive, and communicative; a comparison of these types provides a basis for the selection of home elements. Scenario-oriented design is employed to construct a smart living space (the AI. Living Lab.), which verifies the feasibility of the interface's operations. Finally, the resulting design methods and practical results are used to create a system development blueprint. Real life scenarios provide a basis for communication with smart home industry collaborators, enabling architectural designers to perform scenario modeling of a smart home proposal, quickly and conveniently establish a smart home information modeling system, and promote teamwork with smart home industry personnel.