Obsidian RAGsody

CLI tool for intelligent Obsidian vault interaction using Retrieval-Augmented Generation (RAG).

Features

1. Vault Querying

Ask questions about your vault content using natural language.

• "What did I write about machine learning?"
• "Show me my notes on productivity"

2. URL to Note

Create markdown files from URLs. Files are saved to either the root or to optimal folders based on content similarity.

• "Create a note from https://example.com about AI trends"
• "Summarize this article: https://blog.example.com/post"

Installation

pipx install git+https://github.com/nicolaischneider/obsidianRAGsody.git
obsidian-ragsody

Or run with uv:

uv venv
source .venv/bin/activate
uv sync
uv run python src/main.py

Warning: On first run, you will be asked to enter

• your Obsidian vault path (eg "/Users/your_username/obsidian")
• your OpenAI API key

Tech Stack

• LlamaIndex: RAG framework for vault querying
• OpenAI: GPT-4o-mini for responses, text-embedding-3-small for search
• requests + BeautifulSoup: Web scraping
• rich: Beautiful markdown rendering
• prompt-toolkit: Interactive CLI

Outlook for upcoming features:

• Faster parsing: Upgrade to selectolax for 10x faster HTML parsing
• Index caching: Persistent RAG index storage for faster startup
• Parallel scraping: Concurrent URL processing for multiple links

Requirements

• Python 3.13+
• OpenAI API key
• Obsidian vault (local folder with .md files)
• (optional) pipx for easy CLI installation