There is a significant gap between AI-generated web content and usable products that can support long-term business operations. With the widespread adoption of low-threshold AI development tools, in-depth utilization of these tools will drive new demands, and the complete AI development process is gradually being led by Agents. As one of the earliest AI-native Integrated Development Environments (IDEs) introduced in China, TRAE recently merged TraeWork and TraeCode, and added an Agent Workbench. This improvement not only simplifies operations for beginners but also meets the needs for deep project involvement. TRAE supports scheduling multiple Agents across projects to work in parallel and also allows users to enter IDE mode to view and adjust code. In practical testing, the new version of TRAE built a circular bullet screen display project from scratch, generating an initial solution based on a rich template library and enabling the setup of automated inspection tasks. Project-related documents, code changes, and reports are uniformly retained and circulated, making subsequent iterations smoother. Leveraging the engineering processing capabilities accumulated through its AI-native IDE, TRAE integrates office capabilities into the development process, covering the entire workflow from requirement organization to implementation and operation for the same project. This not only enables more people to develop according to their own work needs but also supports continuous iteration in response to changing requirements. Previously, TRAE has accumulated a large user base, providing a solid foundation for its focus on the professional development track (which can be translated as 'segment' or 'field' depending on context).