by joyehuang
🤖 Official Interactive Tutorial for OpenHarness – Zero to Hero in 12 Chapters | Learn OpenHarness like Claude Code: Agent Loop, Tools, Memory, Multi-Agent | 面向零基础的 AI Agent 交互式教程
Requires a passing catalog security scan. Resolve the flagged issues and resubmit to enable featuring.
# Add to your Claude Code skills
git clone https://github.com/joyehuang/Learn-Open-HarnessGuides for using ai agents skills like Learn-Open-Harness.
Last scanned: 5/29/2026
{
"issues": [
{
"type": "npm-audit",
"message": "brace-expansion: brace-expansion: Large numeric range defeats documented `max` DoS protection",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "express-rate-limit: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "fast-uri: fast-uri vulnerable to path traversal via percent-encoded dot segments",
"severity": "high"
},
{
"type": "npm-audit",
"message": "hono: Hono missing validation of cookie name on write path in setCookie()",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "ip-address: ip-address has XSS in Address6 HTML-emitting methods",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "next: Next.js has a Denial of Service with Server Components",
"severity": "high"
},
{
"type": "npm-audit",
"message": "postcss: PostCSS has XSS via Unescaped </style> in its CSS Stringify Output",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "qs: qs has a remotely triggerable DoS: qs.stringify crashes with TypeError on null/undefined entries in comma-format arrays when encodeValuesOnly is set",
"severity": "medium"
}
],
"status": "WARNING",
"scannedAt": "2026-05-29T07:57:39.309Z",
"semgrepRan": false,
"npmAuditRan": true,
"pipAuditRan": true
}The interactive companion to HKUDS/OpenHarness — built the day after OpenHarness launched.
Zero to Hero in 12 chapters. Learn the architecture behind Claude Code: Agent Loop, Tools, Memory, Permissions, Multi-Agent & more.
Live Demo → · Issues · Contributing · 中文版
Author: joyehuang.me · Xiaohongshu (32k likes) →
⭐ If this tutorial helps you understand AI Agents, star the repo — Goal: 10k stars!
https://github.com/user-attachments/assets/72102803-c56d-4b37-816f-f530a7ce9dc2
OpenHarness dropped and immediately blew up — the first open-source implementation of the AI Agent Harness pattern that powers Claude Code, Cursor, and the next generation of AI coding tools.
But ~11,700 lines of infrastructure code is a lot to digest. Concepts like Agent Loop, tool registries, permission tiers, and MCP aren't obvious from reading source alone.
So I built this the next day: 12 interactive chapters that break down every core concept with real-world analogies, animated diagrams, hands-on simulators, and quizzes — so you can actually understand what OpenHarness does before diving into the code.
No AI or programming experience required.
| 📖 12 Chapters | From "What is an AI Agent" to "Multi-Agent Collaboration" |
| 🎮 Interactive | Agent Loop animation, Tool Explorer, Permission Sandbox, Architecture Diagram |
| 🧠 Quizzes | Embedded in every chapter with instant feedback |
| 📊 Progress | Auto-tracks completed chapters |
| 🌙 Dark Mode | Deep navy palette, easy on the eyes |
| 📱 Responsive | Mobile, tablet, desktop |
Harness = Tools + Knowledge + Observation + Action + Permissions
| # | Chapter | Key Concepts |
|---|---|---|
| 01 | What is AI Agent | LLM vs Agent — from "talking" to "doing" |
| 02 | What is Harness | The Agent's equipment system |
| 03 | The Harness Equation | Tools + Knowledge + Observation + Action + Permissions |
| # | Chapter | Key Concepts |
|---|---|---|
| 04 | Agent Loop | Think → Act → Observe, the infinite loop |
| 05 | Tool System | 43+ built-in tools, unified interface + registry pattern |
| 06 | Permission System | Three-tier security guardrails |
| 07 | Hook System | Lifecycle events, inject custom logic at critical moments |
| # | Chapter | Key Concepts |
|---|---|---|
| 08 | Skills | On-demand knowledge loading |
| 09 | Memory | Cross-session persistence, 4 memory types |
| 10 | MCP Integration | Standardized external world connection (like USB) |
| # | Chapter | Key Concepts |
|---|---|---|
| 11 | Multi-Agent | Parallel sub-agents, divide and conquer |
| 12 | Full Flow | Complete journey from user input to Agent response |
git clone https://github.com/joyehuang/Learn-Open-Harness.git
cd Learn-Open-Harness
npm install
npm run dev
Open http://localhost:3000 and start learning.
| Tech | Purpose |
|---|---|
| Next.js 16 | App Router, SSR |
| TypeScript 5 | Type safety |
| Tailwind CSS v4 | Styling |
| shadcn/ui | UI components |
| next-themes | Dark mode |
Contributions welcome!
src/content/chapter-XX.tssrc/content/src/components/interactive/Each chapter is a TypeScript file with section blocks (text, analogy, code, quiz, key-concept, comparison, diagram) — easy to modify, zero barrier to contribute.
⭐ Star this repo if it helped you — Goal: 10k! ⭐
Built by @joyehuang · Xiaohongshu (32k likes) →
Learn-Open-Harness is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by joyehuang. 🤖 Official Interactive Tutorial for OpenHarness – Zero to Hero in 12 Chapters | Learn OpenHarness like Claude Code: Agent Loop, Tools, Memory, Multi-Agent | 面向零基础的 AI Agent 交互式教程. It has 331 GitHub stars.
Learn-Open-Harness returned warnings in SkillsLLM's automated security scan. It has no critical vulnerabilities, but review the flagged issues in the Security Report section before adding it to your workflow.
Clone the repository with "git clone https://github.com/joyehuang/Learn-Open-Harness" and add it to your Claude Code skills directory (see the Installation section above).
Learn-Open-Harness is primarily written in TypeScript. It is open-source under joyehuang on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh Learn-Open-Harness against similar tools.
No comments yet. Be the first to share your thoughts!