RAG・知識ベース
llm-wiki-agent
A personal knowledge base that builds itself — drop in sources, get a persistent interlinked wiki.
★ 3.6k スターMIT
ガイドを見る →OpenKBは、VectifyAIが公開しているオープンソースのRAG・知識ベースプロジェクトです（GitHub: VectifyAI/OpenKB）。GitHubスターは4,759、フォークは495で、Apache-2.0ライセンスの下で公開されています。GitHubでの最終更新は2026年7月22日です。
OpenKB compiles your documents into a persistent, interlinked wiki knowledge base — no vector DB
Collect the papers, docs, or URLs you want compiled into one knowledge base.
Install OpenKB from PyPI, create a folder, and run the init command, choosing your LLM model.
Add files, a whole directory, or a URL; OpenKB compiles them into wiki pages with cross-links.
Ask questions with cited answers, chat interactively, or generate outputs like skills and slide decks.
Documents become summary, concept, and entity pages with cross-references that accumulate over time; contradictions get flagged.
PageIndex tree indexing handles long PDFs by reasoning over a hierarchical index — no vector database required.
PDF, Word, Markdown, PowerPoint, HTML, Excel, CSV, text, and URLs, with native handling of figures, tables, and images.
One-off cited answers or multi-turn chat over the wiki, with persisted sessions you can resume.
Distill redistributable agent skills from the wiki, visualize an interactive knowledge graph, or generate single-file HTML slide decks.
The wiki is plain Markdown with wikilinks that opens in Obsidian, and pages follow Google's Open Knowledge Format specification.
Clone for a source install, or simply install from PyPI in the next step.
git clone https://github.com/VectifyAI/OpenKB.git
cd OpenKBInstall the package (or `pip install -e .` from the clone for development).
pip install openkbCreate a knowledge base, add a document, and ask your first question (set your LLM API key in .env first).
openkb init
openkb add paper.pdf
openkb query "What are the main findings?"VectifyAI, the team behind PageIndex; OpenKB is powered by PageIndex's open-source vectorless retrieval.
Yes, it is Apache-2.0 licensed. You pay only for the LLM API usage of the provider you configure (OpenAI, Claude, Gemini, and others via LiteLLM); some subscription providers can authenticate via OAuth instead of a key.
It needs an LLM — usually your own API key set in .env. PageIndex runs locally by default; the optional PageIndex Cloud features need a separate key.
RAG rediscovers knowledge from chunks on every query and nothing accumulates. OpenKB compiles knowledge once into a persistent wiki with cross-references already built, then keeps it current as documents are added.
Run `pip install openkb`, then `openkb init` in a fresh folder, `openkb add <file>`, and `openkb query` or `openkb chat`.
A personal knowledge base that builds itself — drop in sources, get a persistent interlinked wiki.
Agent-fed knowledge base that distills learnings into cited wiki pages.
Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.
Turn your GitHub stars into a searchable AI knowledge base.
A knowledge-base platform with visual AI workflows — build Q&A systems on your own documents
Graph RAG conversational knowledge base built on TiDB Serverless vector storage
LibreAIカタログに掲載されたRAG・知識ベースプロジェクトの中で最もスターの多いもの — OpenKBが合わなかった場合に検討したい、現実的な候補リストです。
Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or…
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context…
Lightweight OCR toolkit turning PDFs and images into structured, LLM-ready data in 100+ languages
MinerU — turns PDFs and Office documents into LLM-ready Markdown and JSON for agents and RAG pipelines, Apache-2.0 plus terms
Stop renting your intelligence. Own it with AnythingLLM. Everything you need for a powerful local-first agent experience
カタログの中でOpenKBとスター数が近いRAG・知識ベースプロジェクト — 前後の棚もそのまま見て回れます。
An agent that finds anything on your computer — federated search over your files, in place.
AI-native database for hybrid search across dense, sparse, and tensor vectors plus full text
RAG for your monorepo: query, understand and edit multi-language codebases with a knowledge graph.
RAGApp: deploy a RAG app over your documents with Docker — configure it like a custom GPT
Build Neo4j knowledge graphs from unstructured data using LLMs — then chat with your graph.
Ragent (容纳智能体): a production-grade Java Agentic RAG platform with hybrid retrieval and an admin console