Alibaba Cloud has announced a comprehensive upgrade to its Memory service, with a focus on enhancing long-term memory and personalized processing capabilities for Personal Agent scenarios. The service now offers mainstream access methods such as MCP and CLI. It can extract key information from conversations in real time, generate user profiles and structured memories, and precisely retrieve relevant context during subsequent interactions to reduce repetitive inquiries. In multiple public evaluations, the service's core metrics performed prominently, ranking among the top among participating systems. In practical scenarios, the service can dynamically update user memories and support developers in customizing rules and integrating capabilities into multiple mainstream Agent tool frameworks. It can also work closely with the Skill module to enable efficient collaboration. This upgrade helps personal assistants break through the context bottleneck of single-turn conversations, evolve toward deep personalized long-term companionship, and provides infrastructure support for Agents to delve into complex vertical scenarios.