The complexity of issues confronting landscape architects and other environmental designers, in the face of the global polycrisisvare immense and growing. Artificial Intelligence (AI) tools, may represent a panacea, but they are still very little beyond infancy; results across many domains are so far quite mixed; and limitations and concerns are abundant. In this context, several different but related emergent technological advances do seem likely to be key drivers of the next generation of AIenabled computer-aided-design tools intended to help design thinking and manage real-world complexity: ‘Agentic AI’ (AAI) platforms connected to ‘Digital Twins ’, esp. ‘Digital Twin Earths’ (DTEs), with ‘Model Context Protocol’ (MCP) and ‘Retrieval-Augmented Generation’ (RAG) used to control and constrain the AI agents. Following a brief description of these technologies and their foundations, I explore some application possibilities by re-visiting a decade-old ‘System for Geodesign’ with AAI, MCP, RAG, and DTE integrated, and outline an updated prescription for a technologically enhanced and enabled next-gen ‘AAI+DTE’ design system (not just for geodesign). A highly simplified prototype agentic system implementation is described.