nashsu/llm_wiki is a desktop application that uses an LLM to incrementally build and maintain a persistent, interlinked knowledge wiki from your documents. Unlike traditional RAG that retrieves and answers from scratch each time, it compiles ingested material into an organized structure with source traceability.
Project overview
It addresses a core inefficiency of traditional RAG by compiling documents into a persistent, interlinked wiki with a knowledge graph and community detection, rather than re-deriving answers on every query.
Project type
AI Agent · RAG · Document Processing
Use cases
Knowledge Q&A · Search & Research
Deployment
Refer to project documentation
License
License pending
Best for
Users who want to build and maintain a personal knowledge base from their own documents
Key capabilities
Analyzes documents first, then generates wiki pages with source traceability and incremental cache.
Extracts embedded images from PDFs, generates factual captions with a vision LLM, and surfaces them in search results.
Ingests PDF, Office documents, EPUB/MOBI, Org mode, images, media, web clips, and batches of URLs.
Visualizes relationships using direct links, source overlap, Adamic-Adar, and type affinity, with Louvain community detection.
Provides optional embedding-based retrieval via LanceDB, supporting any OpenAI-compatible endpoint.
Performs multi-query web search via Tavily, SerpApi, or SearXNG, auto-ingesting results into the wiki.
Provides a tool-using chat runtime with wiki, source, graph, web retrieval, workspace file generation, and shell approval.
Offers a built-in local JSON API and bundled MCP server for hybrid search, file read, graph traversal, and source rescan.
Limitations and risks
The application requires an external LLM provider to be configured for operation.
Using paid LLM providers and web search APIs introduces potential costs.
Getting started
Setup is considered easy due to pre-built binaries available for macOS, Windows, and Linux. Users download the binary from Releases, launch the app to create a new project, configure an LLM provider in Settings, import documents via Sources, and use Chat to query the knowledge base.
Evidence and sources
GitHub project description: LLM Wiki is a cross-platform desktop application that turns your documents into an organized, interlinked knowledge base — automatically.
README: - **Two-Step Chain-of-Thought Ingest** — LLM analyzes first, then generates wiki pages with source traceability and incremental cache
README: Instead of traditional RAG (retrieve-and-answer from scratch every time), the LLM **incrementally builds and maintains a persistent wiki** from your sources.
Release: LLM Wiki v0.6.6
README: - **Obsidian compatibility** — the wiki directory works as an Obsidian vault