Overview
Understanding RA-H's approach to knowledge management
What is RA-H?
RA-H is a local-first contextual substrate — it captures your entire intellectual experience into a connected graph on your machine. Reading, ideas, conversations, research — all of it, linked and searchable.
The Problem
You consume more than you can retain. Articles, podcasts, papers, conversations — the good stuff slips through the cracks. Traditional tools ask you to organize upfront. You don't have time for that.
RA-H's Solution
- Captures everything — articles, videos, papers, ideas, conversations. One place.
- Connections emerge — edges carry explanations, not just links. Your graph grows richer over time.
- Search your own thinking — semantic search across everything you've ever captured.
- Local and portable — SQLite on your machine. No cloud. No lock-in. Works with any AI via MCP.
- Structure follows use — dimensions and connections emerge from your content, not predefined folders.
Two Ways to Use RA-H
| Version | Best For | Get It |
|---|---|---|
| Mac App | Most users. One-click install, auto-updates | Download |
| Open Source (RA-OS) | Developers, self-hosters, contributors. BYO API keys | GitHub |
Both versions use the same core codebase. The Mac app adds packaging, auth, and optional subscription features. The open source version is fully functional with your own API keys.
Design Philosophy
Non-prescriptive & Emergent
The system doesn't force you into folders or predefined categories. Organization emerges naturally from your actual content. The structure adapts to how you think, not the other way around.
Everything is Connected
Every piece of knowledge can potentially connect to any other. Connections aren't just links — they carry context, explanation, and meaning. RA-H helps you discover these connections automatically.
Local-First
Your knowledge network belongs to you, not a platform. Your thinking, research, and connections all live on your machine in a portable format you control. No cloud dependencies, no data mining, no vendor lock-in.
Human + AI
You guide, AI assists. Shape behavior with skills and a clean graph. Always in control of your knowledge.
Tech Stack
- Frontend: Next.js 15, TypeScript, Tailwind CSS
- Database: SQLite + sqlite-vec (vector search)
- AI Models: Anthropic Claude + OpenAI GPT via Vercel AI SDK
- Desktop: Tauri (Mac app)
- MCP Server: Local connector for Claude Code and external agents
Current Status
- Version: v0.2.0 (March 2026)
- Platform: Mac (Apple Silicon + Intel), macOS 12+
- Database:
~/Library/Application Support/RA-H/db/rah.sqlite - License: MIT (open source version)
Key Features
- Flexible Pane System — Feed view, Focus panel, collapsible Chat
- Single RA-H Assistant — One always-on runtime grounded in your graph and skills
- Smart Organization — Multi-tag dimensions instead of folders
- Skills — Markdown documents that shape behavior, preferences, and reusable procedures
- Content Extraction — YouTube videos, websites, PDFs automatically processed
- MCP Server — Connect Claude Code and other external agents to your knowledge base