Atomic is a self-hosted, semantically-connected personal knowledge base that transforms markdown notes into an AI-augmented knowledge graph. It targets users who need to organize, synthesize, and query their personal information, offering AI-powered insights and connections between notes. The primary benefit is an intelligent, interconnected system for managing and retrieving knowledge.

How It Works

The core logic resides in the atomic-core Rust crate. Markdown notes ("atoms") are automatically chunked, embedded using vector search via sqlite-vec, and linked based on semantic similarity. This forms a knowledge graph that can be visualized on a spatial canvas, synthesized into wiki articles using LLMs with citations, or queried via an agentic RAG chat interface. It supports pluggable AI providers like OpenRouter (cloud) or Ollama (local).

Quick Start & Requirements

• Primary install/run:
  • Desktop App: Download from GitHub Releases.
  • Docker Compose: git clone https://github.com/kenforthewin/atomic.git, cd atomic, docker compose up -d. Access http://localhost.
  • Standalone Server: cargo run -p atomic-server -- --data-dir ./data serve --port 8080.
• Prerequisites: Node.js 22+, Rust toolchain, platform-specific Tauri v2 dependencies (for desktop). AI provider API keys (OpenRouter) or local Ollama setup.
• Links: GitHub Releases (implied), openrouter.ai, ollama.ai.

Highlighted Details

• Atoms: Markdown notes with hierarchical tagging, source URLs, and automatic chunking.
• Semantic Search: Vector search powered by sqlite-vec.
• Canvas: Force-directed spatial visualization mapping semantic similarity.
• Wiki Synthesis: LLM-generated articles with inline citations from notes.
• Chat: Agentic RAG interface for querying the knowledge base.
• AI Providers: Pluggable support for OpenRouter (cloud) and Ollama (local).
• Browser Extension: Captures web content as atoms.
• MCP Server: Exposes knowledge base to AI tools like Claude.
• iOS App: Native SwiftUI client for mobile access.

Maintenance & Community

No specific details regarding maintainers, community channels (e.g., Discord, Slack), sponsorships, or roadmap are provided in the README.

Licensing & Compatibility

The project is licensed under the MIT License, which is permissive for commercial use and integration into closed-source projects.

Limitations & Caveats

The project relies heavily on external AI services, requiring user configuration and potentially incurring costs (OpenRouter) or local resource usage (Ollama). The browser extension requires manual installation in developer mode. Specific integrations, like the MCP server for Claude, necessitate custom configuration.