iwe vs llm_wiki
Markdown notes as a knowledge graph: an LSP for VS Code, Neovim, Zed and Helix, plus CLI and MCP so AI agents query the same files by structure. Rust, local-first, OKF-compatible. — versus — Desktop app implementing Karpathy's LLM Wiki pattern: an LLM ingests your documents into a persistent, interlinked wiki with a knowledge graph, hybrid search, deep research and an MCP server.
Both give agents a Markdown knowledge graph over MCP; LLM Wiki has an LLM write and maintain the pages, IWE leaves writing to you and gives you and your agent structure-aware tools.
| iwe | llm_wiki | |
|---|---|---|
| Stars | 1.7k | 20k |
| Forks | 90 | 2.3k |
| Language | Rust | TypeScript |
| License | Apache-2.0 | NOASSERTION |
| Last activity | 2 days ago | today |
| Topics | knowledge-graphs, memory | rag, knowledge-graphs |
| Curated connections | 4 | 5 |
iwe — the curator's take
The best fit for people who already live in Markdown and want their agent to share that brain: inclusion links give notes a real hierarchy, the LSP adds rename, refactor and search in your editor, and the CLI and MCP let Claude or Codex navigate by structure (parent context, subtrees, frontmatter filters) instead of similarity guesses. No database, no cloud; git is the sync. IWE has no AI of its own, so it will not write or summarize notes for you (that is LLM Wiki's job), and its payoff scales with how carefully you link. It speaks Google's OKF, so notes stay portable.
llm_wiki — the curator's take
Pick LLM Wiki if you like Karpathy's 'compile knowledge once instead of re-retrieving it every query' idea but don't want to wire it up yourself: drop in PDFs, Office docs, EPUBs or web clips and an LLM writes and maintains source-traced wiki pages, with Louvain clusters, gap-finding and deep research on top, plus an MCP server and a skill so Claude Code or Codex can query it. It's a personal, single-user desktop app — not a team knowledge service (see arkon) — and every ingest spends LLM tokens on a two-step analyse-then-write pass, so big corpora cost real money. Licence is GPL-3.0 (the GitHub API reports NOASSERTION): fine to use, copyleft if you fork and redistribute.