Tencent open-sources a team-level memory hub for AI coding agents
Tencent Cloud just open-sourced TencentDB Agent Memory v2.0, a self-hosted memory layer designed for teams running multiple AI coding agents. The pitch is not just retrieval. It’s governance: controlling which agent can see which memory, and which version of that memory is valid.
Four Memory Asset Types
The system ingests conversations, documents, and code, then organizes them into four distinct asset types:
- Chat Memory: persisted conversation context
- Skill: reusable procedural knowledge extracted from past sessions
- LLM-Wiki: structured factual knowledge your agents can query
- Code-Graph: a graph representation of your codebase for code-aware retrieval
All four are versioned and permissioned at the asset level, not just the database level.
️ The Governance Angle
The differentiator here is ACL-based visibility. Memory Hub narrows by team, user, agent, and visibility scope before it retrieves anything. That means different agents on the same team can operate with different memory views, and you can version-control which knowledge is active without blowing up shared context.
For teams running more than one coding agent in parallel, that kind of isolation matters. Without it, agents pick up each other’s stale or incorrect context and compound the errors.
Integration and Setup
The project is MIT licensed and self-hosted via Docker. It integrates with Claude Code, OpenClaw, Hermes, and CodeBuddy. No SaaS dependency, no vendor lock-in on the memory layer itself.
One caveat from the source: private-repo Code-Graph and automated memory routing are still being refined. Large regulated enterprises should pilot rather than standardize at this point.
I’m founder of BizStack at Cagri Sarigoz LLC and a passionate advocate for entrepreneurs.
With over 14 years in tech, marketing, and AI, including my role as Head of SEO at CitizenShipper and co-founder of TaleBot at Intale AI, I’m dedicated to sharing genuine, useful product insights and tips.
At BizStack, I aim to cut through the digital noise to provide clear, actionable advice.
And more than all else, I’m a father to a (always) little girl and a husband.
Contact me at [email protected] for assistance.
