llm-wiki-compiler
The knowledge compiler. Raw sources in, interlinked wiki out. Inspired by Karpathy's LLM Wiki pattern.
llm-wiki-compiler is a knowledge compilation tool that transforms raw sources—papers, notes, READs, transcripts, PDFs, web pages—into a structured, interlinked markdown wiki with full citation traceability. Implementing Andrej Karpathy's LLM Wiki pattern, it compiles knowledge once into durable pages with typed records, relationships, and provenance, rather than re-discovering context at every query. The tool includes an MCP server that exposes compilation, querying, and knowledge management capabilities to AI agents, plus a TypeScript SDK and local browser viewer. Developers use it to create persistent, auditable knowledge bases that agents and humans can browse, query, lint, and export in multiple formats including Open Knowledge Format (OKF).
Key Features
Use Cases
- 01Compiling research papers, notes, and transcripts into a citation-traceable wiki for scientific projects
- 02Building a codebase documentation knowledge base with typed entities, relationships, and freshness tracking
- 03Creating team handbooks or decision logs with auditable provenance and review workflows
- 04Providing AI agents with stable, citation-aware context packs instead of raw file collections
- 05Managing editorial workflows with typed articles, desks, bylines, and lifecycle states
- 06Exchanging compiled knowledge between tools using portable OKF, JSON-LD, GraphML, or llms.txt exports
Related MCP Servers
View moremarkitdown
Python tool for converting files and office documents to Markdown.
firecrawl
The context API to search, scrape, and interact with the web at scale. 🔥
prompts.chat
f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.
langflow
Langflow is a powerful tool for building and deploying AI-powered agents and workflows.
llm-wiki-compiler — FAQ
What is llm-wiki-compiler?+
llm-wiki-compiler is a knowledge compilation tool that turns raw sources into a structured, citation-traceable markdown wiki with typed pages, relationships, and provenance. It implements Andrej Karpathy's LLM Wiki pattern, compiling knowledge once for reuse rather than re-extracting it at query time.
How do I install llm-wiki-compiler?+
Install globally via npm with `npm install -g llm-wiki-compiler`, then set your provider API key (e.g., `export ANTHROPIC_API_KEY=sk-...`). Node.js 24 or newer is required.
Which AI clients work with llm-wiki-compiler's MCP server?+
The MCP server (`llmwiki serve`) works with any MCP-compatible client, including Claude Desktop, Cursor, and other editors that support the Model Context Protocol. Read-only tools work without provider credentials; LLM-backed tools require provider configuration at call time.
Do I need an API key to use llm-wiki-compiler?+
Yes, for LLM-backed compilation and querying. The default provider is Anthropic (requires `ANTHROPIC_API_KEY`), but you can use OpenAI, Ollama, GitHub Copilot, Atlas Cloud, or other compatible providers. Some read-only operations like viewing and lint checks work without credentials.
Is llm-wiki-compiler free and open source?+
Yes, llm-wiki-compiler is open source under the MIT license. However, you will incur costs from your chosen LLM provider (Anthropic, OpenAI, etc.) when running compilation and query operations.
What is the difference between llm-wiki-compiler and traditional RAG?+
Traditional RAG retrieves raw chunks and reconstructs relationships at every query. llm-wiki-compiler compiles sources once into typed, interlinked pages with citations and metadata, then queries operate over that stable artifact. This makes knowledge durable, auditable, and reusable across queries.
How do I install llm-wiki-compiler?+
Open the source repository on GitHub and follow its README. llm-wiki-compiler is a mcp server — MCP Agents Market links you directly to the official repo.
Is llm-wiki-compiler free?+
llm-wiki-compiler is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.