@DanKornas: Your Markdown notes are useful until the answer spans five files. Kwipu is a local Graph RAG system for people who keep…

X AI KOLs Timeline Tools
graph-rag markdown obsidian local-llm mcp-server open-source knowledge-management

Summary

Kwipu is a new open-source local Graph RAG system that indexes Markdown and Obsidian notes into a property graph, enabling natural-language queries with hybrid retrieval and supporting local LLMs via Ollama.

Your Markdown notes are useful until the answer spans five files. Kwipu is a local Graph RAG system for people who keep project knowledge in Markdown or Obsidian vaults. It helps you ask natural-language questions across scattered notes by turning Markdown files into a persisted property graph and querying it with hybrid retrieval. Key features: • Markdown-to-graph indexing – builds a property graph from notes and LLM-extracted relationships • Obsidian-aware parsing – reads [[wikilinks]] and YAML frontmatter into structured triples • Hybrid retrieval – combines optional synonym expansion, vector similarity, BM25, and temporal/metadata matching • Local Ollama workflow – runs LLM and embedding models locally, with CLI model overrides • MCP server – lets Claude Desktop, Cursor, Windsurf, or other MCP-compatible agents query the graph It’s open-source (MIT license). Link in the reply
Original Article
View Cached Full Text

Cached at: 06/17/26, 01:45 AM

Your Markdown notes are useful until the answer spans five files.

Kwipu is a local Graph RAG system for people who keep project knowledge in Markdown or Obsidian vaults.

It helps you ask natural-language questions across scattered notes by turning Markdown files into a persisted property graph and querying it with hybrid retrieval.

Key features: • Markdown-to-graph indexing – builds a property graph from notes and LLM-extracted relationships • Obsidian-aware parsing – reads [[wikilinks]] and YAML frontmatter into structured triples • Hybrid retrieval – combines optional synonym expansion, vector similarity, BM25, and temporal/metadata matching • Local Ollama workflow – runs LLM and embedding models locally, with CLI model overrides • MCP server – lets Claude Desktop, Cursor, Windsurf, or other MCP-compatible agents query the graph

It’s open-source (MIT license).

Link in the reply

Similar Articles

@neil_xbt: https://x.com/neil_xbt/status/2067083332513395140

X AI KOLs Timeline

This article explains how to build a knowledge graph in Obsidian using Claude as an AI engine to find connections, arguing that note-taking systems become more valuable over time when notes are linked, rather than isolated.