ai-agent

llm-wiki-compiler: Compile Knowledge

Name: llm-wiki-compiler: Compile Knowledge
Author: atomicstrata

Get an interlinked wiki from raw sources with llm-wiki-compiler, used by founders

1,842 stars179 forksTypeScriptGuide quality 8/10Updated 7/28/2026100% free · open source

What it does

The llm-wiki-compiler takes raw sources of information and compiles them into an interlinked wiki, streamlining the process of creating a knowledge base.

When to use it

•When you have a large amount of unorganized knowledge that needs to be structured and linked together
•When creating a wiki or documentation for a project and want to automate the process
•When integrating Karpathy's LLM Wiki pattern into your workflow and need a tool to compile the data

Ready-to-paste prompt

node dist/index.js --input sources.json --output wiki

Heads up: Make sure you have Node.js and TypeScript installed on your system, as well as the necessary dependencies, which can be installed by running 'npm install' in the project directory

Saves to your device

Use with Claude

New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

🛠️ Technical setup

Expect 20–40 minutes in a terminal — or let your AI agent drive it.

Try it instantly — no install

Claude Code

mkdir -p ~/.claude/skills/llm-wiki-compiler && curl -fsSL https://workflowstacks.com/api/skills/llm-wiki-compiler/claude-skill -o ~/.claude/skills/llm-wiki-compiler/SKILL.md

Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)

claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

How llm-wiki-compiler: Compile Knowledge works

Codeflow

Free to inspect

Llm-wiki-compiler: Compile Knowledge is a very large TypeScript project (~92k lines across 498 code files, plus 763 test files). It is a full software project: use it through its install path rather than reading it end to end. Last commit this month, MIT license, has a test suite.

Size

Very large codebase

~92k lines · 498 code files · days to read — use, don't read

Setup

Developer setup

A real software project. Use it via its install path; don't expect to read it all.

Runs on

Node.js

No API keys detected

TypeScript 95%JavaScript 4%CSS 1%

Where to start reading

1
README.md
Start here — what it does and how to install it
2
AGENTS.md
The instructions the AI actually follows
3
src/cli.ts
Where the program starts running
4
package.json
Dependencies and the commands it exposes
5
examples/basic/README.md
A worked example — copy this to get going

What's in each folder

src/Core code — the actual logic507 files

docs/Documentation57 files

examples/Examples you can copy10 files

scripts/Helper scripts6 files

test/Tests — proof it works799 files

.github/CI / automation (GitHub Actions)2 files

.husky/Tool / agent settings2 files

READMEHas testsDocumentedCI checksExamples includedMIT licenseUpdated this month

Use with Claude

New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

🛠️ Technical setup

Expect 20–40 minutes in a terminal — or let your AI agent drive it.

Try it instantly — no install

Claude Code

mkdir -p ~/.claude/skills/llm-wiki-compiler && curl -fsSL https://workflowstacks.com/api/skills/llm-wiki-compiler/claude-skill -o ~/.claude/skills/llm-wiki-compiler/SKILL.md

Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)

claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

Quick Actions

Details

Creator

atomicstrata

Language

TypeScript