Featured

Deploy OpenClaw in 60 seconds — 20% off logoDeploy OpenClaw in 60 seconds — 20% off

Launch OpenClaw on Hostinger in about 60 seconds and keep your agent live 24/7. Our referral link gives you 20% off, no coupon code needed.

Launch on Hostinger
Run your Hermes agent on Hostinger, fully managed logoRun your Hermes agent on Hostinger, fully managed

Launch Hermes on Hostinger in one click, fully managed, no VPS knowledge needed. Use code ZACAARON10 for 10% off.

Launch on Hostinger
Crawl and scrape any site into clean data, 10% off logoCrawl and scrape any site into clean data, 10% off

Firecrawl crawls and scrapes any site into clean markdown for your agent. Get 1,000 free credits, and new users get 10% off their first purchase.

Try Firecrawl free
6,000+ web scrapers for your AI agent, start free logo6,000+ web scrapers for your AI agent, start free

Apify gives your agent live web data: 6,000+ prebuilt scrapers and actors, MCP-ready. Sign up free with $5 in usage credits.

Try Apify free
One API to scrape, enrich, and extract the internet. logoOne API to scrape, enrich, and extract the internet.

Context.dev gives your agents a single API to scrape, enrich, and extract live web data — no proxies, no parsers, no maintenance.

Start building free
SetupClaw: done-for-you OpenClaw for founders & exec teams logoSetupClaw: done-for-you OpenClaw for founders & exec teams

White-glove OpenClaw for founders and exec teams (4–50+ employees): we install, harden, integrate your tools, and maintain it — secured from day one.

Get it set up for you
SEO data APIs for your agent, $1 free credit logoSEO data APIs for your agent, $1 free credit

DataForSEO gives your agent live access to SERP results, keyword data, backlinks, and on-page SEO data through one API. New accounts get a $1 credit, good for up to 20,000 keyword or backlink lookups.

Try DataForSEO free
Reach 48,000+ AI builders

A flat monthly placement in front of developers actively installing AI tools. No lock-in, cancel anytime.

Advertise here

Works with

Claude CodeClaude DesktopCursorVS CodeClineCodex CLIOpenClaw+ any MCP client

Install to Claude Code

This server doesn't publish a one-line install command. Follow the setup in the source repository.

Summary

MCP server for persistent, compounding markdown wikis maintained by LLMs. Enables incremental knowledge base building with interlinked pages, search, and raw source management.

README.md

LLM Wiki MCP

MCP server for LLM-maintained markdown wikis — a persistent, compounding knowledge base instead of per-query RAG.

Inspired by Andrej Karpathy's LLM Wiki pattern. The LLM incrementally builds and maintains interlinked markdown pages; you curate sources and ask questions; the wiki stays current because maintenance cost is near zero.

Three layers

| Layer | Path | Who writes | |---|---|---| | Raw sources | raw/ | You (immutable) | | Wiki | wiki/{type}/{slug}.md | LLM via MCP | | Schema | AGENTS.md | You + LLM |

Plus index.md (catalog) and log.md (timeline), maintained automatically on writes.

MCP tools

| Tool | Description | |---|---| | wiki_get_index | Read master index.md | | wiki_list_pages | List pages, optional type filter | | wiki_get_page | Fetch one page (frontmatter + body) | | wiki_search | Full-text search with snippets | | wiki_list_raw | List immutable source files in raw/ | | wiki_read_raw | Read a raw source file (text or binary metadata) | | wiki_put_page | Create/update page + optional log/index/git commit |

Quick start

1. Create your wiki from the starter template

cp -r examples/starter-wiki ~/my-wiki
cd ~/my-wiki
git init
git add . && git commit -m "Initial wiki scaffold"

2. Configure Cursor

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "llm-wiki": {
      "command": "npx",
      "args": ["-y", "@jlnkrth/llm-wiki-mcp"],
      "env": { "WIKI_DIR": "/path/to/my-wiki" }
    }
  }
}

Restart Cursor. Copy examples/starter-wiki/AGENTS.md conventions into your wiki (included if you copied the starter).

3. Local development (from source)

git clone https://github.com/jlnkrth/llm-wiki-mcp.git
cd llm-wiki-mcp
npm install
WIKI_DIR=/path/to/my-wiki npm run start:stdio

See TESTING.md for the full test checklist.

Remote deployment (HTTP + Docker)

For team/shared wikis on a VPS:

cd deploy
cp .env.example .env
docker compose up -d --build

MCP endpoint: http://<host>/wiki-mcp/mcp with header X-Api-Key: <MCP_API_KEY>.

See deploy/README.md for details.

Configuration

wiki.config.yaml in your wiki root:

types:
  - entities
  - concepts
  - sources
  - synthesis
  - topics
git:
  enabled: true    # false = local writes only, no commit/push
  branch: main
  commit_prefix: "wiki:"
search:
  refresh_interval_ms: 300000

Roadmap

See ROADMAP.md for planned features (local image assets, rich query outputs, search upgrades).

Why this works

RAG rediscovers knowledge on every question. An LLM wiki compiles once and stays current — cross-references exist, contradictions are flagged, synthesis reflects everything you've read. The MCP is the safe write interface: single-writer lock, index/log maintenance, structured frontmatter.

Browse the wiki in Obsidian; let the agent maintain it.

License

MIT

See related servers & alternatives →

Related MCP servers

Browse all →

Related guides

Hand-picked reading to help you choose and use Vector & Memory servers.