LLM Wiki

Build and maintain a persistent, compounding knowledge base where the LLM does all the grunt work — summarizing, cross-referencing, filing, and bookkeeping — while you focus on sourcing, exploration, and asking the right questions.

Core Principle: Stop using LLMs like search engines. Use them as knowledge compilers. Instead of re-deriving knowledge from raw documents on every query (RAG), the LLM incrementally builds a persistent wiki — structured, interlinked markdown files that compound in value with every source you add and every question you ask. "Humans abandon wikis because the maintenance burden grows faster than the value. LLMs don't get bored." — Karpathy

Prerequisites

• No runtime dependencies (methodology-only skill, no scripts)
• Any LLM Agent with file read/write capability (Claude Code, Cursor, Codex, OpenCode, etc.)
• A file system (local or remote) for storing markdown files
• Optional: Obsidian for browsing/graph view, Git for version control

When to Use

Invoke when: