repowiserepowise

tencentcloud / tencentdb-agent-memory Architecture and code health

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.

★ 30,956stars on GitHubTypeScript

This connects Repowise to tencentcloud/tencentdb-agent-memory's code. It does not install tencentdb-agent-memory.

Repowise maps public repos from their source. Not affiliated with tencentcloud. What is Repowise?

github.com/tencentcloud/tencentdb-agent-memoryIndexed at 42d4cde

What is tencentdb-agent-memory?

tencentdb-agent-memory is a team-level memory hub for AI Agents, meant to stop project context, documents, code and working methods from being re-explained in every new session. It turns agent conversations, docs and code into four reusable memory assets: Chat Memory, Skill, LLM-Wiki and Code-Graph. These assets can be governed, shared and equipped across agents and frameworks. The project is written mainly in TypeScript. To install it, run git clone https://github.com/tencentcloud/tencentdb-agent-memory.git and follow the steps below. Its code health is 6.8 out of 10 (Fair), ranked #164 of 4,000+ indexed repos. The last commit to tencentdb-agent-memory was on Nov 9, 2026, and 37 people have contributed.

As of commit 42d4cde, indexed Nov 10, 2026. What it is comes from a model reading the README; the rest is measured from the source and git history, not from the maintainer.

For agentstencentdb-agent-memory.md

Repo facts: type, license, health, rank, contributors
tencentdb-agent-memory facts
Code health6.8 / 10 (Fair)
Health rank#164 of 4,000+ indexed repos
Last commitNov 9, 2026
Contributors37
Bus factor1
Size1,203 files, 196,543 lines

Install tencentdb-agent-memory

Commands copied from the README

  1. 1Clone the tencentdb-agent-memory repository.
    git clone https://github.com/tencentcloud/tencentdb-agent-memory.git
  2. 2Change into the deployment directory of the repository.
    cd TencentDB-Agent-Memory/deploy/global-images
  3. 3Create the environment file from the example and fill in the two sets of LLM parameters in it.
    cp .env.example .env

Releases on GitHub (latest v2.0.4, Nov 19, 2026)

Full install notes on GitHub

Explore tencentdb-agent-memory

Ask about this code in your agent

This connects Repowise to tencentdb-agent-memory's code. It does not install tencentdb-agent-memory. To install tencentdb-agent-memory, use the install steps.

Useful if you want to change or fork tencentdb-agent-memory: your agent can ask how it is put together and get answers that cite its files.

Ask about tencentdb-agent-memory

Answers come from tencentdb-agent-memory's README and source, with the files they cite. Two free answers a day without an account.

Featured answer

tencentdb-agent-memory is a team-level memory hub for AI Agents that turns agent conversations, docs and code into four reusable memory assets: Chat Memory, Skill, LLM-Wiki and Code-Graph. Its goal is to keep project context, documents, code and working methods from being re-explained in every new session, and these assets can be governed, shared and equipped across agents and frameworks. Inputs enter through MemoryProxy, which handles session initialization, session recovery and per-request metadata, and through MemoryCore, whose entry point is MemoryCore/index.ts and which builds and persists memory artifacts. MemoryKnowledge exposes HTTP routes and wiki and code-graph services that index those artifacts, and the MemoryPanel web UI shows the resulting chat memory, skills, and wiki and code-graph pages to team managers. The deploy/global-images directory holds start-all.sh, which launches the memory, hub and proxy services together. The codebase has 1203 files, with the largest share in the Application layer (685 files) and the Service layer (149 files), so start with MemoryCore/index.ts and then the layers listed by file count.

Files per layer in tencentdb-agent-memory10 layers.

Basis: Inferred · knowledge graph· as of 42d4cde

Want this for your own codebase? Map your repo free

About this page

Get a page like this for your own codebase

Repowise built this map of tencentcloud/tencentdb-agent-memory from its source and git history. It parses every symbol, draws the dependency graph, scores each file's health from complexity, duplication, coverage and churn, finds hotspots and owners in the history, and writes documentation for every module. Your coding agent reads the same index over MCP, so its answers cite real files.

Free for 2 public repos, with a page like this one. No card. See how Repowise works

Prefer local? pip install repowise && repowise init indexes your own checkout, no account needed. Star Repowise on GitHub

tencentcloud/tencentdb-agent-memory FAQ

What is tencentdb-agent-memory?

tencentdb-agent-memory is a team-level memory hub for AI Agents that turns conversations, docs and code into reusable memory assets. The assets are Chat Memory, Skill, LLM-Wiki and Code-Graph, and they can be shared across agents and frameworks.

How does tencentdb-agent-memory work?

tencentdb-agent-memory receives agent conversations and tool calls through a proxy, then builds and stores memory artifacts in its MemoryCore pipeline. Agents receive that context back through the proxy, and the MemoryPanel web UI lets teams review and share the assets.

How do I install tencentdb-agent-memory?

tencentdb-agent-memory is installed from its deploy/global-images directory by copying the environment example to .env, filling in the LLM parameters, and running start-all.sh. That script launches the memory, hub and proxy services together. See the install steps

What language is tencentdb-agent-memory written in?

tencentdb-agent-memory is written primarily in TypeScript. Its repository also contains shell, Python, JavaScript and Dockerfile files.

How is the tencentdb-agent-memory code organized?

The tencentdb-agent-memory code is split into the MemoryCore, MemoryProxy, MemoryPanel and MemoryKnowledge modules, plus sdk, deploy, agents and adapters directories. MemoryCore is the largest module by file count.

When should I use tencentdb-agent-memory?

tencentdb-agent-memory is for teams whose Agents keep repeating project context, document reading and workflows. It lets new sessions reuse existing chat memory, skills, wiki pages and code graphs.

Code health

6.7out of 10Fair

Hotspots average 4.5, the place fixes pay off most.

On the leaderboards:#54 of 70 TypeScript repos

Key features of tencentdb-agent-memory

  • Chat Memory keeps preferences, facts, decisions and interaction history, and session bindings and permissions decide which memories an Agent can access.
  • Skills are extracted from conversations and tool calls, and each Skill has versions, resource files, trigger boundaries, execution steps and validation rules.
  • Wiki turns product docs, design specs and runbooks into structured pages with a link graph, and supports Markdown, plain text, HTML, Word, Excel, PowerPoint and PDF.
  • CodeGraph indexes code symbols, files, call relationships and impact paths so Agents can inspect callers and callees before changing code.
  • Memory assets can be private, shared with the whole team, or restricted to specific Users, Roles or Agents through access control lists.

How tencentcloud/tencentdb-agent-memory works

Agent conversations, tool calls, documents and code enter the system through the MemoryProxy and MemoryCore parts. MemoryProxy handles session initialization, session recovery and per-request metadata, so requests reach the right agent and task context. MemoryCore/index.ts is the entry point for the MemoryCore pipeline, which builds and persists memory artifacts, and MemoryKnowledge exposes HTTP routes and wiki and code-graph services that index those artifacts. The MemoryPanel web UI shows the resulting chat memory, skills, and wiki and code-graph pages to people who manage teams and assets. The deploy/global-images directory contains the start-all.sh script that launches the memory, hub and proxy services together.

The codebase has 1,203 files and 12,056 symbols in 8 modules, led by TypeScript, Shell and Python. Code health is 6.7 out of 10, rated fair.