LLMWiki

A blazing fast CLI and TUI for maintaining an LLM-managed personal knowledge base. Built in Rust, powered by veclite-db, and integrated locally with Ollama.

Setup & Installation

You will need Rust installed on your machine to build this project.

Install directly from GitHub using Cargo:

cargo install --git https://github.com/rithulkamesh/llmwiki

Or clone and build from source:

git clone https://github.com/rithulkamesh/llmwiki
cd llmwiki
cargo install --path .

1. Make sure you have Ollama running with an embedding model (e.g. nomic-embed-text) and a generation model (e.g. gemma4 or llama3).

ollama run gemma4
ollama pull nomic-embed-text

2. Initialize your wiki folder:

mkdir ~/notes && cd ~/notes
llmwiki init

Features

• llmwiki: Boot the interactive TUI dashboard.
• llmwiki sync: Crawl .md files, chunk, embed, and index them into veclite.db concurrently.
• llmwiki ask <query>: RAG semantic synthesis. Searches your notes and answers with gemma4.
• llmwiki ingest <url>: CLI web-clipper. Downloads, extracts, auto-summarizes, and saves the URL as a new markdown page.
• llmwiki lint: Check your entire vault for broken links and orphan pages.

Config

Your settings are saved in .llmwiki.yaml in your wiki directory.

ollama_url: "http://192.168.1.2:11434"
model_name: "gemma4"
embed_model: "nomic-embed-text"