liangdabiao

基于 [Karpathy llm-wiki]方法论,利用 AI 持续构建和维护你的个人知识库。支持从多种素材源(网页、推特、公众号、小红书、知乎、YouTube、PDF、本地文件)自动整理为结构化的 wiki,并通过 Quartz 发布为静态wiki知识库网站。 并通过 claude_agent_sdk 调用 claude agent 使用 llm-wiki ,提供api接口对外访问服务 。

14
3
89% credibility
Found Apr 30, 2026 at 14 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

A system for building and maintaining a personal AI-powered structured wiki from notes, PDFs, and web content using agentic workflows and static site generation.

How It Works

1
🔍 Discover llm-wiki

You hear about a simple way to turn your notes, PDFs, and web links into a smart, connected personal wiki using AI.

2
📥 Get and set up

Download the tool and run one command to create your wiki folder with everything ready to go.

3
📝 Add your stuff

Drop in links to articles, PDFs, or notes, and the AI starts organizing them automatically.

4
🤖 AI builds your wiki

Watch as the AI reads everything, creates linked pages, summaries, and connections between ideas.

5
👀 Preview your site

Click to build and see your beautiful, searchable knowledge base come alive on a simple website.

Your wiki grows

Now you have a living knowledge hub that gets smarter as you add more, easy to browse and share.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 14 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is llm-wiki?

This TypeScript project, inspired by Karpathy's llm-wiki gist and github repo like karpathy/nanogpt or autoresearch, automates building a personal knowledge base from sources like webpages, Twitter, WeChat articles, Xiaohongshu, Zhihu, YouTube, PDFs, and local files. AI (via Claude agents) structures content into a wiki, publishes it as a Quartz static site for easy hosting, and exposes an API for querying. Run agentic scripts to ingest and generate pages without manual workflows.

Why is it gaining traction?

It stands out with seamless multi-source ingestion and agentic Claude skills for autonomous wiki maintenance, echoing Karpathy's llm-wiki twitter ideas and openclaw skills. Developers dig the zero-config Quartz output for fast static deploys and API endpoints for integration, skipping clunky manual curation in tools like Obsidian plugins or basic llm wiki 2 setups.

Who should use this?

AI enthusiasts or researchers mirroring Karpathy's llm-wiki graphify approach for personal wikis. Knowledge workers scraping Twitter/YouTube for structured notes, or Claude users needing API-driven access to auto-maintained sites without full RAG pipelines.

Verdict

Solid early experiment at 14 stars—high 0.8999999761581421% credibility score, but light docs and tests mean tweak for production. Try if you're into Karpathy llm-wiki prompts; fork for custom skills.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.