Lapis0x0/obsidian-yolo

📌 What is YOLO?

YOLO ("Agent‑native AI assistant for Obsidian") is a community plugin that brings large‑language‑model agents directly into the Obsidian note‑taking app. It lets you chat with, write, and orchestrate AI tools without leaving your vault, turning your personal knowledge base into an interactive, RAG‑enabled workspace.

🚀 Core Capabilities (as described in the README)

Capability	What it does
Agent Experience	Connects to locally‑signed‑in CLI agents (Claude Code, Codex, Hermes, Pi, Grok) and lets them operate on your vault – calling tools, running shell commands, performing web searches, etc.
Vault RAG	Indexes your Obsidian notes (including local embedding models) and retrieves context‑grounded answers, so the AI can cite your own content.
Learning Mode	Turns any topic + source material into a structured study project: outlines, knowledge points, flashcards, spaced‑repetition (FSRS) and an Anki `.apkg` export.
Multi‑Window Chat	Run several independent chat sessions in parallel, each with its own context and task.
Memory System	Persists preferences, habits, and long‑term context across conversations.
Cursor Chat & Quick Ask	One‑click `@` in the editor to ask a question, edit, or continue writing without leaving the note.
Tab Completion	Real‑time AI‑driven suggestions while you type.
Multi‑Model Support	Switch among OpenAI, Anthropic Claude, Google Gemini, DeepSeek, Groq, etc.
External Agent / MCP Support	Hook up external agents (e.g., Hermes, OpenClaw) or delegate tasks to a configured YOLO Agent.
Whiteboard	A high‑performance Obsidian Canvas that works faster than the native canvas, intended for brainstorming with AI assistance.
i18n	Full multilingual UI.

🛠️ Getting Started (quick‑start steps from the README)

Install – Open Obsidian → Settings → Community plugins → Browse → search "YOLO" → install & enable.
Configure – Add an API key for the model you want (OpenAI, Anthropic, Gemini, Grok, etc.) or use an OAuth sign‑in.
Start chatting – Open the YOLO sidebar, or type @ in any note to launch Quick Ask.
(Optional) Manual install – Download main.js, manifest.json, styles.css from the latest release, place them in <vault>/.obsidian/plugins/obsidian-yolo/, then enable.

Note: YOLO conflicts with the Smart Composer plugin; you must disable the latter first.

📚 Documentation & Help

Full user guide: documentation/en/README.md
Topics include Getting Started, Model setup, Chat workflow, Sparkle (quick‑ask & tab completion), Knowledge Base, Tools & Permissions, Memory, Skills, Agents, MCP integration, CLI agents, and more.
Community support via a Discord server (link in README) and issue tracker on GitHub.

🎯 Who Might Benefit?

Knowledge workers who keep extensive notes in Obsidian and want AI‑augmented search and writing.
Students & lifelong learners looking for a personal study system with flashcards and spaced‑repetition.
Developers who already use CLI AI agents and want to bring that power into their note‑taking workflow.
Teams that need a unified AI assistant that can call external tools, run shell commands, or perform web searches from within their shared vault.

🧩 How It Fits Into the AI Landscape

RAG (Retrieval‑Augmented Generation) – YOLO indexes your vault locally (or via remote embeddings) and retrieves relevant passages to ground model responses.
Agent orchestration – Supports tool‑calling, sub‑agents, and MCP (Multi‑Channel Protocol) servers, turning a simple Q&A bot into an active collaborator.
Multi‑model flexibility – Users can swap between major LLM providers without leaving Obsidian.

📦 License & Contributions

Licensed under the MIT License.
Contributions are welcomed; the repo includes a CONTRIBUTING.md with guidelines, an AI‑assisted PR policy, and a request‑first workflow for major features.

TL;DR: YOLO is a full‑featured AI‑agent plugin for Obsidian that adds chat, RAG, multi‑model support, tool calling, and a learning system, turning your note vault into an interactive AI‑powered workspace.