GitHub 上一个叫《Learn Harness Engineering》的开源课程最近上了 Hacker News 首页（155 分、5.1k stars）， 主题是 AI coding agent 的可靠性工程——也就是 OpenAI 和 Anthropic 最近几个月一直在讲的 harness engineering。

https://github.com/walkinglabs/learn-harness-engineering…

作者是清华 SIGS 的开源团队 walkinglabs，课程内容 12 节 lecture + 6 个实战项目，全部围绕同一个压轴项目（一个 Electron 知识库 app），目前已经有 13 种语言版本。

核心讲的是这几条：

1、Harness Engineering 是新概念但绝不只是营销： OpenAI 2026 年 2 月在 Codex 官方博客正式提出，Anthropic 同期发了两篇配套工程文章。这门课是把这两家分散的工程内容系统整理成中文可读的体系，不是发明新东西；

2、核心论点是模型不是问题，环境才是： Anthropic 自己跑过一个对照实验，同样的 Opus 4.5、同样的提示词（“做一个 2D retro 游戏编辑器”），没 harness 花 $9 跑 20 分钟出不来；有完整 harness（planner + generator + evaluator）花 $200 跑 6 小时跑出来一个能玩的游戏。差距不在模型，在环境；

3、Harness 的 5 个子系统： Instructions（agent 看什么）、State（已经做了什么）、Verification（怎么算完成）、Scope（一次只做一个 feature）、Lifecycle（会话起手和收尾）；

4、最小起手 4 个文件： AGENTS.md（操作手册）、http://init.sh（每次起手跑的环境检查脚本）、feature_list.json（机器可读的功能边界）、claude-progress.md（会话之间的进度交接）。这 4 个文件直接丢进你现有代码库就能用，不用读完 12 节课；

5、为什么 agent 总是搞砸（课程反复回答的几个问题）： 一次想做 3 件事但 3 件都只做一半；说“done“但其实 test 跑不过；跨会话完全没有记忆，这次写到一半下次重头来；没有验证流水线就没有真完成（只有 test pass + lint pass + e2e pass 三件齐了才算）；

6、6 个项目渐进式演进： P01 prompt-only vs rules-first 对照看差距、P02 把代码库改成 agent 能读的结构、P03 跨会话续接、P04 加运行时反馈控制范围、P05 让 agent 自验、P06 完整 harness 加可观测性；

7、压轴项目是个 Electron 知识库 app： 每个项目的答案就是下一个项目的起手代码——app 在演进，你的 harness 技能也在演进；

总结：

“AI agent 工程师“这个角色正在从“会写提示词的人“变成“会搭 harness 的人”。

这门课本身不算硬核，但它把 OpenAI 和 Anthropic 两家工程博客分散的内容系统整理成了中文索引，对中文圈想入门这个方向的人来说，能省下读 5 篇英文工程博客的时间。

walkinglabs/learn-harness-engineering

Source: https://github.com/walkinglabs/learn-harness-engineering

Learn Harness Engineering

A project-based course on building the environment, state management, verification, and control mechanisms that make AI coding agents work reliably.

Globe icon This course is available in 13 languages: English, 简体中文, 繁體中文, 日本語, 한국어, Español, Français, Русский, Deutsch, العربية, Tiếng Việt, Oʻzbekcha, Türkçe. Choose your language from the badges above.

Learn Harness Engineering is a course dedicated to the engineering of AI coding agents. We have deeply studied and synthesized the most advanced Harness Engineering theories and practices in the industry. Our core references include:

• OpenAI: Harness engineering: leveraging Codex in an agent-first world
• Anthropic: Effective harnesses for long-running agents
• Anthropic: Harness design for long-running application development
• Awesome Harness Engineering

Quick start? The skills/harness-creator/ skill can help you scaffold a production-grade harness (AGENTS.md, feature lists, init.sh, verification workflows) for your own project in minutes.

Table of Contents

• ✨ Visual Preview
• What Harness Engineering Actually Means
• Quick Start: Improve Your Agent Today
• Capstone Project: A Real App
• Learning Path
• Syllabus
• Skills
• Other Courses

✨ Visual Preview

🏠 Course Homepage

A comprehensive course outline and introduction to core philosophies, providing a clear path to get started.

📖 Immersive Lectures

Deep dives into real-world pain points and hands-on projects (like Project 01) for an immersive learning experience.

🗂️ Ready-to-Use Resource Library

Templates and reference configurations designed to solve common pitfalls in multi-turn AI agent development, such as context loss and premature task completion.

PDF Coursebooks

The repository now includes a PDF build pipeline for the course content.

• Run npm run pdf:build to generate the currently configured PDF coursebooks locally.
• Output files are written to artifacts/pdfs/.
• Run npm run screenshots:readme if you want to refresh the README preview images.
• GitHub Actions workflow release-course-pdfs.yml can build the PDFs and publish them to GitHub Releases.

The Model Is Smart, The Harness Makes It Reliable

There’s a hard truth most people learn the hard way: the strongest model in the world will still fail on real engineering tasks if you don’t build a proper environment around it.

You’ve probably seen this yourself. You give Claude or GPT a task in your repo. It starts well — reads files, writes code, looks productive. Then something goes wrong. It skips a step. It breaks a test. It says “done” but nothing actually works. You spend more time cleaning up than if you’d done it yourself.

This isn’t a model problem. It’s a harness problem.

The evidence is clear. Anthropic ran a controlled experiment: same model (Opus 4.5), same prompt (“build a 2D retro game editor”). Without a harness, it spent $9 in 20 minutes and produced something that didn’t work. With a full harness (planner + generator + evaluator), it spent $200 in 6 hours and built a game you could actually play. The model didn’t change. The harness did.

OpenAI reported the same thing with Codex: in a well-harnessed repository, the same model goes from “unreliable” to “reliable.” Not a marginal improvement — a qualitative shift.

This course teaches you how to build that environment.

                    THE HARNESS PATTERN
                    ====================

    You --> give task --> Agent reads harness files --> Agent executes
                                                        |
                                              harness governs every step:
                                              |
                                              +--> Instructions: what to do, in what order
                                              +--> Scope:       one feature at a time, no overreach
                                              +--> State:       progress log, feature list, git history
                                              +--> Verification: tests, lint, type-check, smoke runs
                                              +--> Lifecycle:   init at start, clean state at end
                                              |
                                              v
                                         Agent stops only when
                                         verification passes

What Harness Engineering Actually Means

Harness engineering is about building a complete working environment around the model so it produces reliable results. It’s not about writing better prompts. It’s about designing the system the model operates inside.

A harness has five subsystems:

    ┌─────────────────────────────────────────────────────────────────┐
    │                        THE HARNESS                              │
    │                                                                 │
    │   ┌──────────────┐  ┌──────────────┐  ┌──────────────────────┐ │
    │   │ Instructions  │  │    State     │  │   Verification       │ │
    │   │              │  │              │  │                      │ │
    │   │ AGENTS.md    │  │ progress.md  │  │ tests + lint         │ │
    │   │ CLAUDE.md    │  │ feature_list │  │ type-check           │ │
    │   │ feature_list │  │ git log      │  │ smoke runs           │ │
    │   │ docs/        │  │ session hand │  │ e2e pipeline         │ │
    │   └──────────────┘  └──────────────┘  └──────────────────────┘ │
    │                                                                 │
    │   ┌──────────────┐  ┌──────────────────────────────────────┐   │
    │   │    Scope     │  │         Session Lifecycle             │   │
    │   │              │  │                                      │   │
    │   │ one feature  │  │ init.sh at start                     │   │
    │   │ at a time   │  │ clean-state checklist at end          │   │
    │   │ definition   │  │ handoff note for next session        │   │
    │   │ of done      │  │ commit only when safe to resume      │   │
    │   └──────────────┘  └──────────────────────────────────────┘   │
    │                                                                 │
    └─────────────────────────────────────────────────────────────────┘

    The MODEL decides what code to write.
    The HARNESS governs when, where, and how it writes it.
    The harness doesn't make the model smarter.
    It makes the model's output reliable.

Each subsystem has one job:

• Instructions — Tell the agent what to do, in what order, and what to read before starting. Not one giant file; a progressive disclosure structure the agent navigates on demand.
• State — Track what’s been done, what’s in progress, and what’s next. Persisted to disk so the next session picks up exactly where the last one left off.
• Verification — Only a passing test suite counts as evidence. The agent cannot declare victory without runnable proof.
• Scope — Constrain the agent to one feature at a time. No overreach. No half-finishing three things. No rewriting the feature list to hide unfinished work.
• Session Lifecycle — Initialize at the start. Clean up at the end. Leave a clean restart path for the next session.

Why This Course Exists

The question isn’t “can models write code?” They can. The question is: can they reliably complete real engineering tasks inside real repositories, over multiple sessions, without constant human supervision?

Right now, the answer is: not without a harness.

    WITHOUT HARNESS                          WITH HARNESS
    ==============                          ============

    Session 1: agent writes code            Session 1: agent reads instructions
              agent breaks tests                      agent runs init.sh
              agent says "done"                       agent works on one feature
              you fix it manually                     agent verifies before claiming done
                                                       agent updates progress log
    Session 2: agent starts fresh                    agent commits clean state
              agent has no memory
              of what happened before         Session 2: agent reads progress log
              agent re-does work                       agent picks up exactly where it left off
              or does something else entirely          agent continues the unfinished feature
              you fix it again                         you review, not rescue

    Result: you spend more time                  Result: agent does the work,
            cleaning up than if you                      you verify the result
            did it yourself

The questions this course actually cares about:

• Which harness designs improve task completion rates?
• Which designs reduce rework and incorrect completions?
• Which mechanisms keep long-running tasks progressing steadily?
• Which structures keep the system maintainable after multiple agent runs?

Course Curriculum & Documentation

For the full course materials, please visit the Documentation Website.

The curriculum is divided into three parts:

1. Lectures: 12 conceptual units explaining the theory behind harness engineering.
2. Projects: 6 hands-on projects where you build an agentic workspace from scratch.
3. Resource Library: Copy-ready templates (AGENTS.md, feature_list.json, init.sh, etc.) to use in your own repositories today.

Quick Start: Improve Your Agent Today

You don’t need to read all 12 lectures before you start getting value. If you’re already using a coding agent on a real project, here’s how to improve it right now.

The idea is simple: instead of just writing prompts, give your agent a set of structured files that define what to do, what’s been done, and how to verify the work. These files live inside your repo, so every session starts from the same state.

    YOUR PROJECT ROOT
    ├── AGENTS.md              <-- the agent's operating manual
    ├── CLAUDE.md              <-- (alternative, if using Claude Code)
    ├── init.sh                <-- runs install + verify + start
    ├── feature_list.json      <-- what features exist, which are done
    ├── claude-progress.md     <-- what happened each session
    └── src/                   <-- your actual code

Grab the starter templates from the Resource Library and drop them into your project. That’s it. Four files, and your agent sessions will already be significantly more stable than running on prompts alone.

Capstone Project: A Real App

All six course projects revolve around the same product: an Electron-based personal knowledge base desktop app.

    ┌─────────────────────────────────────────────────────┐
    │               Knowledge Base Desktop App            │
    │                                                     │
    │  ┌──────────────┐  ┌──────────────────────────────┐│
    │  │ Document List │  │       Q&A Panel              ││
    │  │              │  │                              ││
    │  │ doc-001.md   │  │  Q: What is harness eng?    ││
    │  │ doc-002.md   │  │  A: The environment built    ││
    │  │ doc-003.md   │  │     around an agent model... ││
    │  │ ...          │  │     [citation: doc-002.md]   ││
    │  └──────────────┘  └──────────────────────────────┘│
    │                                                     │
    │  ┌─────────────────────────────────────────────────┐│
    │  │ Status Bar: 42 docs | 38 indexed | last sync 3m ││
    │  └─────────────────────────────────────────────────┘│
    └─────────────────────────────────────────────────────┘

    Core features:
    ├── Import local documents
    ├── Manage a document library
    ├── Process and index documents
    ├── Run AI-powered Q&A over imported content
    └── Return grounded answers with citations

This project was chosen because it combines strong practical value, enough real-world product complexity, and a good setting for observing before/after harness improvements.

Each course project’s starter/solution is a complete copy of this Electron app at that evolutionary stage. P(N+1)’s starter is derived from P(N)’s solution — the app evolves as your harness skills grow.

Learning Path

The course is designed to be done in order. Each phase builds on the last.

    Phase 1: SEE THE PROBLEM              Phase 2: STRUCTURE THE REPO
    ========================              ==========================

    L01  Strong models ≠ reliable         L03  Repository as single
         execution                              source of truth
    L02  What harness actually means
                                       L04  Split instructions across
         |                                   files, not one giant file
         v
    P01  Prompt-only vs.                       |
         rules-first comparison                v
                                               P02  Agent-readable workspace


    Phase 3: CONNECT SESSIONS             Phase 4: FEEDBACK & SCOPE
    ==========================           =========================

    L05  Keep context alive               L07  Draw clear task boundaries
         across sessions
                                       L08  Feature lists as harness
    L06  Initialize before every               primitives
         agent session
                                               |
         |                                     v
         v                                     P04  Runtime feedback to
    P03  Multi-session continuity                   correct agent behavior


    Phase 5: VERIFICATION                 Phase 6: PUT IT ALL TOGETHER
    =====================                 ============================

    L09  Stop agents from                 L11  Make agent's runtime
         declaring victory early               observable

    L10  Full-pipeline run =              L12  Clean handoff at end of
         real verification                      every session

         |                                     |
         v                                     v
    P05  Agent verifies its own work       P06  Build a complete harness
                                               (capstone project)

Each phase takes about a week if you’re going part-time. If you want to go faster, phases 1–3 can be done in a long weekend.

Syllabus

Lectures — 12 conceptual units, each answering one core question

Read the full text for each lecture on the Documentation Website.

Projects — 6 hands-on projects applying lecture methods to the same Electron app

    PROJECT EVOLUTION
    =================

    P01  Prompt-only vs. rules-first       You see the problem
     |
     v
    P02  Agent-readable workspace           You restructure the repo
     |
     v
    P03  Multi-session continuity           You connect sessions
     |
     v
    P04  Runtime feedback & scope           You add feedback loops
     |
     v
    P05  Self-verification                  You make the agent check itself
     |
     v
    P06  Complete harness (capstone)        You build the full system

    Each project's solution becomes the next project's starter.
    The app evolves. Your harness skills grow with it.

Resource Library

• English — templates, checklists, and method references
• 简体中文 — 中文模板、清单和方法参考
• 繁體中文 — 繁體中文範本、清單和方法參考
• 日本語 — テンプレート、チェックリスト、方法リファレンス
• 한국어 — 템플릿, 체크리스트, 방법 참고 자료
• Español — plantillas, listas de verificación y referencias
• Français — modèles, listes de contrôle et références
• Русский — шаблоны, чек-листы и справочники
• Deutsch — Vorlagen, Checklisten und Referenzen
• العربية — قوالب، قوائم تحقق ومراجع
• Tiếng Việt — mẫu, danh sách kiểm tra và tài liệu tham khảo
• Oʻzbekcha — andozalar, tekshiruv roʻyxatlari va maʼlumotnomalar

The Agent Session Lifecycle

One of the core ideas in this course: the agent’s session should follow a structured lifecycle, not a free-for-all. Here’s what that looks like:

    AGENT SESSION LIFECYCLE
    ======================

    ┌──────────────────────────────────────────────────────────────────┐
    │  START                                                          │
    │                                                                  │
    │  1. Agent reads AGENTS.md / CLAUDE.md                           │
    │  2. Agent runs init.sh (install, verify, health check)          │
    │  3. Agent reads claude-progress.md (what happened last time)    │
    │  4. Agent reads feature_list.json (what's done, what's next)    │
    │  5. Agent checks git log (recent changes)                       │
    │                                                                  │
    │  SELECT                                                          │
    │                                                                  │
    │  6. Agent picks exactly ONE unfinished feature                   │
    │  7. Agent works only on that feature                             │
    │                                                                  │
    │  EXECUTE                                                         │
    │                                                                  │
    │  8. Agent implements the feature                                 │
    │  9. Agent runs verification (tests, lint, type-check)           │
    │  10. If verification fails: fix and re-run                      │
    │  11. If verification passes: record evidence                    │
    │                                                                  │
    │  WRAP UP                                                         │
    │                                                                  │
    │  12. Agent updates claude-progress.md                           │
    │  13. Agent updates feature_list.json                            │
    │  14. Agent records what's still broken or unverified            │
    │  15. Agent commits (only when safe to resume)                   │
    │  16. Agent leaves clean restart path for next session           │
    │                                                                  │
    └──────────────────────────────────────────────────────────────────┘

    The harness governs every transition in this lifecycle.
    The model decides what code to write at each step.
    Without the harness, step 9 becomes "agent says it looks fine."
    With the harness, step 9 is "tests pass, lint is clean, types check."

Who This Is For

This course is for:

• Engineers already using coding agents who want better stability and quality
• Researchers or builders who want a systematic understanding of harness design
• Tech leads who need to understand how environment design affects agent performance

This course is not for:

• People looking for a zero-code AI introduction
• People who only care about prompts and don’t plan to build real implementations
• Learners not prepared to let agents work inside real repositories

Requirements

This is a course where you actually run coding agents.

You need at least one of these tools:

• Claude Code
• Codex
• Another IDE or CLI coding agent that supports file editing, command execution, and multi-step tasks

The course assumes you can:

• Open a local repository
• Allow the agent to edit files
• Allow the agent to run commands
• Inspect output and re-run tasks

If you don’t have such a tool, you can still read the course content, but you won’t be able to complete the projects as intended.

Local Preview

This repository uses VitePress as a documentation viewer.

npm install
npm run docs:dev        # Dev server with hot reload
npm run docs:build      # Production build
npm run docs:preview    # Preview built site

Then open the local URL that VitePress outputs in your browser.

Prerequisites

Required:

• Familiarity with the terminal, git, and local development environments
• Ability to read and write code in at least one common application stack
• Basic software debugging experience (reading logs, tests, and runtime behavior)
• Enough time to commit to implementation-focused coursework

Helpful but not required:

• Experience with Electron, desktop apps, or local-first tools
• Background in testing, logging, or software architecture
• Prior exposure to Codex, Claude Code, or similar coding agents

Core References

Primary:

• OpenAI: Harness engineering: leveraging Codex in an agent-first world
• Anthropic: Effective harnesses for long-running agents
• Anthropic: Harness design for long-running application development
• OpenAI: Unrolling the Codex agent loop
• Anthropic: Demystifying evals for AI agents
• LangChain: Improving Deep Agents with harness engineering
• Thoughtworks / Martin Fowler: Harness engineering for coding agent users
• Cursor: Continually improving our agent harness

See the full layered reference list in docs/en/resources/reference/.

Repository Structure

learn-harness-engineering/
├── docs/                          # VitePress documentation site
│   ├── lectures/                  # 12 lectures (index.md + code/ examples)
│   │   ├── lecture-01-*/
│   │   ├── lecture-02-*/
│   │   └── ... (12 total)
│   ├── projects/                  # 6 project descriptions
│   │   ├── project-01-*/
│   │   └── ... (6 total)
│   └── resources/                 # Multilingual templates & references
│       ├── en/                    # English templates, checklists, guides
│       ├── zh/                    # Chinese templates, checklists, guides
│       ├── ru/                    # Russian templates, checklists, guides
│       └── vi/                    # Vietnamese templates, checklists, guides
├── projects/
│   ├── shared/                    # Shared Electron + TypeScript + React foundation
│   └── project-NN/               # Per-project starter/ and solution/ directories
├── skills/                        # Reusable AI agent skills
│   └── harness-creator/           # Harness engineering skill
├── package.json                   # VitePress + dev tooling
└── CLAUDE.md                      # Claude Code instructions for this repo

How the Course Is Organized

• Each lecture focuses on one question
• The course includes 6 projects
• Every project requires the agent to do real work
• Every project compares weak vs. strong harness results
• What matters is the measured difference, not how many docs were written

Skills

This repository also includes reusable AI agent skills that you can install directly into your IDE or agent workspace.

• harness-creator: A skill that helps you scaffold a production-grade harness for your own project in minutes.

Other Courses

Our team has also created other courses! Check them out:

Hands-on Modern RL: An open-source, hands-on curriculum bridging the gap from basic RL concepts to LLM alignment, RLVR, and advanced Agentic systems.

Star History

Acknowledgments

This course was inspired by and draws ideas from learn-claude-code — a progressive guide to building an agent from scratch, from a single loop to isolated autonomous execution.