cabinet vs TencentDB-Agent-Memory
Self-hosted AI knowledge and team memory compared
cabinet and TencentDB-Agent-Memory both help teams preserve context for AI agents in self-hosted systems. cabinet centers on a Markdown knowledge base and operational workflows, while TencentDB-Agent-Memory organizes reusable memory assets for sharing across agents.
cabinet: Build a Self-Hosted AI Knowledge Base
Cabinet combines a local, Markdown-based knowledge base with AI agents, scheduled jobs, and team workflows. It is aimed at people who want persistent AI context and automation while keeping their files on disk and under their control.
TencentDB-Agent-Memory: Share Reusable Memory Across AI Agents
TencentDB Agent Memory is a team memory hub that turns conversations, documents, and code into reusable assets for AI agents. It suits teams that want agents to share curated context and skills across sessions and frameworks.
| cabinet | TencentDB-Agent-Memory | |
|---|---|---|
| Language | TypeScript | TypeScript |
| License | MIT | NOASSERTION |
| Stars | 2.9k | 27.7k |
| Forks | 302 | 2.7k |
| Last analyzed | Oct 3, 2026 | Oct 3, 2026 |
Key differences
- cabinet stores knowledge as Markdown files on disk with Git-backed history; TencentDB-Agent-Memory distills conversations, documents, and code into managed memory assets.
- cabinet includes goals, missions, tasks, Kanban boards, scheduled jobs, and embedded HTML apps; TencentDB-Agent-Memory focuses on Chat Memory, Skills, Wiki, CodeGraph, and agent asset management.
- cabinet supports local Claude Code and Codex CLI adapters; TencentDB-Agent-Memory provides a proxy integration for supported agent clients and agent bindings for assets.
- cabinet requires Node.js 22 or later and at least one supported AI CLI provider; TencentDB-Agent-Memory requires Node.js 22.16 or newer, configures LLM parameters for memory and proxy groups, and documents Docker deployment.
- cabinet is listed with an MIT license; TencentDB-Agent-Memory is reported by GitHub as NOASSERTION, while its README displays an MIT badge.
Choose cabinet if you…
- want Markdown files on disk with Git-backed history and portable project context.
- need scheduled agent jobs, task workflows, or a local workspace with embedded HTML apps.
- already use Claude Code or Codex CLI and are comfortable running a local web application.
Choose TencentDB-Agent-Memory if you…
- need to share curated memory assets across multiple agents or team members.
- want to manage Skills, linked documentation, code relationships, and agent access controls in one hub.
- plan to seed team memory by importing repositories, documents, or conversation sessions.
This comparison is generated with AI from the OSRepos analyses of both projects. Always check each project's repository and documentation before choosing.