Repolode
Submit
VectifyAI

VectifyAI/OpenKB

A CLI that compiles raw documents into an interlinked, LLM-maintained wiki knowledge base, using vectorless reasoning-based retrieval for long documents.

Open-source alternative to:NotebookLM

View on GitHubWebsite
Stars
4,790
Forks
497
Open issues
83
Last commit
3 months ago

What is it?

OpenKB (Open Knowledge Base) is a command-line system that takes raw files such as PDFs, Word documents, slides, spreadsheets, web pages and URLs, and uses an LLM to turn them into a structured wiki of summaries, concept pages, entity pages and cross-links. When a new document is added, the LLM updates existing pages rather than leaving each file isolated, so knowledge accumulates across the collection instead of being re-derived on every query.

Short documents are converted to Markdown and read in full. Long PDFs are indexed by PageIndex into a hierarchical tree that the LLM navigates, which avoids a traditional vector database. Queries and multi-turn chat run over the compiled wiki with cited answers, and generators can produce agent skills, interactive knowledge graphs or HTML slide decks. The wiki is plain Markdown, so it opens in Obsidian. Models are accessed through LiteLLM, so OpenAI, Claude, Gemini and other providers are supported.

Key features

  • 01Converts PDF, Word, PowerPoint, Excel, HTML, CSV, text and URLs into a compiled wiki.
  • 02Indexes long PDFs with PageIndex tree indexing instead of a vector database.
  • 03Retrieves figures, tables and images alongside text.
  • 04Answers one-off questions with citations or runs multi-turn chat with saved sessions.
  • 05Provides a bundled web UI and a REST API.
  • 06Produces plain Markdown output compatible with Obsidian and agent tools like Claude Code.

Use cases

  • Building a personal or team knowledge base from a growing set of research papers.
  • Querying a large collection of reports or manuals with cited answers.
  • Turning a document collection into a browsable Obsidian graph of linked concepts.
  • Distilling a body of reading into a reusable agent skill for coding assistants.

Topics

#llm#knowledge-base#rag#cli#wiki#document-retrieval#obsidian#agents#ai#retrieval

FAQ

What is OpenKB?
A CLI that compiles raw documents into an interlinked, LLM-maintained wiki knowledge base, using vectorless reasoning-based retrieval for long documents. OpenKB (Open Knowledge Base) is a command-line system that takes raw files such as PDFs, Word documents, slides, spreadsheets, web pages and URLs, and uses an LLM to turn them into a structured wiki of summaries, concept pages, entity pages and cross-links. When a new document is added, the LLM updates existing pages rather than leaving each file isolated, so knowledge accumulates across the collection instead of being re-derived on every query.
Is OpenKB free to use? What license does it have?
OpenKB is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is OpenKB still maintained?
Updated within 6 months. The last commit to OpenKB was on Jul 22, 2026.
What can I use OpenKB for?
Building a personal or team knowledge base from a growing set of research papers. Querying a large collection of reports or manuals with cited answers. Turning a document collection into a browsable Obsidian graph of linked concepts. Distilling a body of reading into a reusable agent skill for coding assistants.
Where can I get OpenKB?
The source code and installation instructions are on GitHub: https://github.com/VectifyAI/OpenKB

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects