On August 6, Tencent Cloud announced the release of Agent Memory version 2.0.0, a major update that introduces Team Memory—an enhanced feature that extends long-term memory capabilities from individual users to team collaboration scenarios.

With this upgrade, code knowledge, project documents, historical conversations, and work methods generated during team-agent interactions can now be stored as shared long-term memory. This memory can be accessed across different agents and dynamically assembled based on each agent's role and specific task requirements, eliminating the need to repeatedly reintroduce project context.

Tencent Cloud stated that users with existing assets such as GitHub code repositories, project documents, or historical agent sessions don't need to start from scratch. Team Memory supports direct import of these existing resources and automatically generates corresponding Wiki entries, CodeGraphs, Chat Memories, and Skills, enabling agents and team members to quickly take over ongoing projects.

Since its open-source release in May, TencentDB Agent Memory has attracted significant developer interest. Within 80 days of going open source, the project surpassed 15,000 stars on GitHub and has repeatedly topped the GitHub Trending daily chart.