TencentCloud/TencentDB-Agent-Memory

TencentDB Agent Memory is a team-level memory hub for AI Agents — turning conversations, docs, and code into four reusable memory assets (Chat Memory, Skill, LLM-Wiki, Code-Graph) that are governed, shared, and equipped across agents and frameworks.

What it solves

It addresses the problem of repetitive work and "cold starts" when using AI agents. Instead of re-explaining project context, re-reading documents, or rediscovering workflows in every new session, this project allows agents to save, organize, and reuse experience across different sessions, frameworks, and team members.

How it works

The system acts as a centralized memory server that agents connect to via a proxy. It transforms raw data into structured "Memory Assets":

  • Chat Memory: Distills raw conversations into layers (L0 raw $\rightarrow$ L1 atoms $\rightarrow$ L2 scenarios $\rightarrow$ L3 personas) to retain preferences and facts.
  • Skills: Extracts reusable executable workflows (including versions and validation rules) from successful tasks.
  • Wiki: Converts documentation into structured pages with a link graph.
  • CodeGraph: Indexes code symbols and call relationships to allow agents to perform impact analysis.

These assets are managed through a Memory Hub control panel where humans can review, version, and assign specific "loadouts" of memory to different agents based on their roles and access permissions (Private, Team, or Restricted).

Who it’s for

  • Developers and teams building multi-agent systems who want to share knowledge across different agents.
  • Users of AI coding agents (like Claude Code or DeepSeek Harness) who want their agents to remember project-specific constraints and history.
  • Small teams or "one-person companies" wanting to build a growing library of agent expertise.

Highlights

  • Zero-code integration: Uses a proxy to point an agent's base URL to the memory server without needing plugins or MCP servers.
  • Layered Memory: Moves from raw logs to high-level personas to prevent context window overflow.
  • Role-based Loadouts: Allows assigning different memory assets to different agents (e.g., a "Reviewer" agent gets the release checklist skill).
  • Cold Start Support: Ability to import existing codebases, documents, and conversation histories to immediately equip a new agent team.

Related

  • Project
    mem9-ai/mem9

    A persistent memory layer for AI agents that enables shared context and recall across sessions, machines, and different agent runtimes.

  • Project
    beilusaiying/always-accompany

    A multi-purpose AI agent framework focused on context and attention mechanisms, providing layered memory and dynamic retrieval for long-term companionship, coding, and work.

  • Project
    Fergana-Labs/stash

    A persistent knowledge base and shared memory layer for AI agents that connects to various data sources and records session transcripts to maintain context across runs.

  • Project
    doobidoo/mcp-memory-service

    A persistent, shared memory backend for AI agents that provides a knowledge graph and semantic search across sessions and tools via REST API and MCP.

  • Project
    peterfei/ai-agent-team

    A virtual AI development team with specialized roles (PM, Backend, Frontend, etc.) and a persistent memory system for managing complex software projects.