Yidu Zhilun has established its presence on Tencent WorkBuddy, marking the launch of its professional medical evidence-based capabilities.

On September 2, Tencent WorkBuddy's open platform officially launched, covering more than 20 fields including healthcare and finance. In the first phase, it opens five major ecological capabilities: Buddy applications, experts, Skills, connectors, and hardware. YIDU TECH (02158)'s clinical evidence-based intelligent agent "Yidu Zhixun" Skill is among the first to settle in the WorkBuddy medical AI workstation. YIDU TECH and Tencent had previously reached a multi-dimensional strategic cooperation, and this represents a further deepening of their ecological collaboration. With WorkBuddy serving as a universal AI work entry point, Yidu Zhixun's professional evidence-based medical capabilities can reach doctors and professional users in a more lightweight Skill form, opening new pathways for evidence-based capabilities into more AI task scenarios. From an independent product to a professional Skill, enabling evidence-based capabilities to be callable medical functions. Agents are changing the way people interact with AI. In the past, users had to actively enter each independent application to find the required capabilities; in the Agent era, tasks can be disassembled, and different Skills can be called upon as needed to complete them. For the highly specialized healthcare industry, where there are clear boundaries of responsibility and high professional thresholds, it is necessary for well-constructed and validated vertical capabilities to participate. Yidu Zhixun's Skill entering the WorkBuddy medical AI workstation provides precisely the professional evidence-based medical capabilities. Yidu Zhixun is an evidence-based intelligent agent platform created by YIDU TECH, built on years of medical data governance and AI capability accumulation, aimed at fulfilling doctors' clinical diagnosis and research needs. It integrates core technical capabilities such as large model semantic understanding, medical knowledge graphs, and evidence-based knowledge retrieval and generation (RAG), selecting a vast number of clinical guidelines, high-quality medical literature, and drug knowledge databases to establish a complete integrated intelligent decision support system of "problem understanding evidence retrieval reasoning generation result traceability." It also refers to international evidence grading systems like GRADE for evidence classification, achieving "traceability of every statement and verifiability of findings," addressing the practical pain points of doctors regarding "difficulty in finding evidence, slow literature reading, and high costs of cross-language research." Yidu Zhixun now launches in a lightweight Skill plugin form on the WorkBuddy medical AI workstation, allowing healthcare professionals to quickly activate it on the WorkBuddy platform without complicated local deployment. It facilitates tasks such as medical evidence retrieval, guideline interpretation, diagnostic assistance, and risk assessment, turning clinical-grade evidence-based capabilities into readily available smart office tools. Callable professional capabilities are opening up the medical AI ecosystem. The initial entry of Yidu Zhixun's Skill into the Tencent WorkBuddy medical AI workstation is significant not just for adding a distribution channel but for further standardizing and modularizing long-accumulated evidence-based capabilities so that they can be called upon by different platforms, Agents, and workflows as needed. Throughout this year, the company has continued to advance the modularization and standardization of medical intelligent agent capabilities. The AI middle platform is now compatible with two new standards: MCP model contextual protocols and Skill modularized skill paradigms, and supports the combination of medical knowledge, large model capabilities, and professional tools to quickly build medical intelligent agents across different specialties and scenarios. Currently, YIDU TECH has collaborated with medical institutions and experts to create over 280 intelligent agent scenarios covering clinical documentation generation, diagnostic assistance, risk assessment, patient education, and more. Within the hospital, Yidu Zhixun can integrate with clinical applications like the intelligent clinical assistant Copilot, embedding itself into doctors' workstations and treatment processes; outside the hospital, Yidu Zhixun directly serves doctors through Apps and Mini Programs. Now, with the entry of the Zhixun Skill into the WorkBuddy medical AI workstation, its professional capabilities are further extending into the open Agent ecosystem. From products and platforms to Skills, the change is in how professional capabilities are called and distributed. Medical AI is moving from "users actively seeking an application" to gradually transitioning to "Agents calling upon suitable professional capabilities based on tasks"; what remains unchanged is the high demand of medical scenarios for professionalism, credibility, and verifiability. In the future, medical AI may not have just one "super entry point," but rather an extensive array of medical Skills that have undergone professional construction and real-world scenario validation, flexibly collaborating across different Agents, terminals, and workflows. For YIDU TECH, this entry is also a further exploration of a method for capability opennesstransforming long-accumulated medical knowledge, evidence-based capabilities, and scenario experiences into more standardized, reusable capability modules, allowing reliable medical intelligence to access more entry points and serve more professional scenarios. When professional medical capabilities can be safely and accurately summoned, medical AI will advance from merely "providing answers" to a new stage of "supporting the completion of professional tasks."