2O

2233admin/obsidian-llm-wiki

开发工具

本页由 AI 翻译,技术细节请以原文为准。 查看原文

38 stars 13 forks 质量 60 趋势 32GPL-3.0

你的Markdown保险库,编译成一个6人格MCP团队,涵盖Claude Code、Codex、OpenCode和Gemini CLI。无头优先。引用,不猜测。

概览

: 英文(本页) ·西班牙语 ·简体中文 — : 英文 ·简体中文 — : 主页 ·建筑 ·理由 ·常见问题解答 你正在阅读这篇文章,因为你的团队已经失去了知识。 不是因为没人写下来。他们写了:文件、会议记录、收回结果、截图、代理回复。问题更糟:这些知识没有状态。没有来源。没有评审。没有晋升路径。无法区分选秀和团队真相。 LLM Wiki 给这堆乱七八糟的修改器加了编译器: 将原始材料保存。将其整理成wiki/摘要、概念页、反向链接和矛盾报告。向代理提问。将有用的答案归档到00-Inbox/AI-Output/。仅将经过审查的知识推广到决策、架构和运行手册中。 它不是AI伴侣。它是一个经过审核的团队内存编译器。Obsidian是IDE,Git/Gitea的审查是账本,MCP/CLI工具是执行表面。 灵感来自Andrej Karpathy的LLM维基。

概述(AI 翻译)

LLM Wiki 将团队的原始研究文件夹变成一个可审核、可查询、自我改进的 Obsidian 维基。无头优先。引用,不猜测。

语言:英语(本页)·西班牙语 ·简体中文 — Guide:英语 ·简体中文 — Wiki:主页 ·建筑 ·理由 ·常见问题解答

你正在阅读这篇文章,因为你的团队已经失去了知识。

不是因为没人写下来。他们写了:文件、会议记录、收回结果、截图、代理回复。问题更糟:这些知识没有状态。没有来源。没有评审。没有晋升路径。无法区分选秀和团队真相。

LLM Wiki 给这堆乱七八糟的修改器加了编译器:

将源材料放入raw/。将其汇编成wiki/摘要、概念页、反向链接和矛盾报告。向引用的代理提问。将有用的答案归档到00-Inbox/AI-Output/。仅将已审查的知识推广到决策、架构和运行手册中。

它不是AI伴侣。它是一个经过审核的团队内存编译器。Obsidian是IDE,Git/Gitea的审查是账本,MCP/CLI工具是执行表面。

灵感来自Andrej Karpathy的LLM维基。

本页由 AI 翻译,技术细节请以原文为准。

README

以下为原文 README 节选,完整内容请查看原文。

LLM Wiki

LLM Wiki turns a team’s raw research folder into a reviewed, queryable, self-improving Obsidian wiki. Headless-first. Cites, doesn’t guess.

Language: English (this page) · Español · 简体中文 — Guide: English · 简体中文 — Wiki: Home · Architecture · Rationale · FAQ

You are reading this because your team has already lost knowledge.

Not because nobody wrote it down. They did: papers, meeting notes, repo findings, screenshots, agent answers. The problem is worse: the knowledge has no state. No source. No reviewer. No promotion path. No way to tell a draft from team truth.

LLM Wiki gives that mess a compiler pass:

capture -> compile -> ask -> file -> review -> promote

Put source material in raw/. Compile it into wiki/ summaries, concept pages, backlinks, and contradiction reports. Ask agents cited questions. File useful answers into 00-Inbox/AI-Output/. Promote only reviewed knowledge into decisions, architecture, and runbooks.

It is not an AI companion. It is a reviewed team memory compiler. Obsidian is the IDE, Git/Gitea review is the ledger, and MCP/CLI tools are the execution surface.

Inspired by Andrej Karpathy’s LLM Wiki. Markdown is the source of truth; the compiler turns structure into a graph; MCP exposes it.


Quick start (Claude Code)

Inside any Claude Code session:

/plugin marketplace add 2233admin/obsidian-llm-wiki
/plugin install llmwiki@obsidian-llm-wiki

Start Claude Code inside your Obsidian vault and LLM Wiki uses it; no clone, no build, no paths to type. To use one vault from anywhere, enter its path when Claude Code asks (or later with /plugin configure llmwiki@obsidian-llm-wiki). You need Node.js 20.18.1+. The plugin ships the MCP server, all /llmwiki:vault-* knowledge-work roles, and the thinking/research commands.

Run /llmwiki:doctor first: it checks the setup, offers a starter layout for an empty vault, and suggests what to ask.

Update with /plugin marketplace update obsidian-llm-wiki, or turn on auto-update for the marketplace in /plugin. Change the vault with /plugin configure llmwiki@obsidian-llm-wiki.

在 GitHub 查看完整 README

安装

This server does not publish a one-line install command.

Open the repository installation guide 新手?先看 mcpskill.org 教程 →