In the AI era, everything runs as an agent. Yao gives you the cage — you decide what goes inside. This tutorial series takes you from a minimal two-file agent to a multi-phase autonomous system, step by step. ## What You'll Learn This series covers five chapters: | Chapter | Topic | What You'll Build | |---------|-------|-------------------| | **Getting Started** (this chapter) | First agent + execution model | A working agent in under 10 minutes | | **Yao Agent** | Hooks, MCP tools, Memory, multi-agent | A bookmark assistant with tools and memory | | **CLI Agent** | Sandbox, SKILL ecosystem, file workspace | An autonomous coding agent | | **Advanced** | Private database + SUI pages | Agent with its own data and UI | | **Robot** | Multi-phase autonomous pipeline | A scheduled, event-driven worker | ## Who This Is For - **Developers** with basic programming experience (TypeScript knowledge helps) - **Coding AI Agents** (Cursor, Claude Code) — see the [SKILL package](../skill/SKILL.md) for a machine-readable version of this knowledge ## How to Read This Pick the path that fits you: **I want to get something running quickly** → Read this chapter top to bottom, then jump to [Yao Agent](/tutorials/agent/yao-agent) **I understand the basics, I want to build a real project** → Skim this chapter, then go straight to [CLI Agent](/tutorials/agent/cli-agent) **I want autonomous, scheduled agents** → Complete chapters 2–3 first, then read [Robot](/tutorials/agent/robot) ## This Chapter Two pages, 15 minutes total: 1. **[Your First Agent](./first-agent)** — Create a minimal agent with just a prompt. Get it running and talking. 2. **[How It Works](./how-it-works)** — Understand the project structure, the three execution modes, and which path to take next.