Awesome LLM Wiki

A curated list of foundational blueprints, functional frameworks, and technical guides for building compounding, AI-compiled knowledge bases.

598 resources in 14 sections and 28 subcategories. Press / to search from any page, or browse by category.

What is an LLM Wiki?

An LLM Wiki is a knowledge base that an AI agent writes and maintains for you as plain, interlinked Markdown files. The pattern was introduced by Andrej Karpathy as an alternative to retrieving document fragments at question time.

  • Compiled, not retrieved

    Instead of searching raw document chunks on every question (RAG), the agent reads each source once and compiles what it learns into lasting wiki pages.

  • Plain files you own

    Raw sources, the generated wiki and a schema of rules live as ordinary files, readable in any editor and versionable with Git.

  • Kept up by the agent

    Ingesting sources, answering queries and linting for stale or contradictory pages are routine jobs the agent runs, so the knowledge compounds over time.

Inspired by a paradigm shift in software development engineering, this architecture treats large language models not as ephemeral chat engines, but as stateful knowledge compilers. Instead of parsing fragmented document partitions dynamically at runtime via traditional RAG loops, these systems leverage autonomous agents to read static source files and systematically construct an interlinked, persistent Markdown knowledge topology.

New to the idea? Start with Foundations.

Browse the collection

Every section and subcategory of the list, in the order the README keeps them.

Know something that belongs here?

The list is curated on GitHub. Read the short contribution rules, then open a pull request against the README.