mirror of
https://github.com/walkinglabs/learn-harness-engineering.git
synced 2026-09-19 05:06:27 +08:00
sanbuphy
a1db89298f
新增 Lecture 13（从手动驱动到自动循环）和 Project 07（搭建你的第一个自动循环）。 核心内容： - L13: 从 /goal 到六原语，generator/evaluator 分离，四种沉默成本，从零搭建第一个 loop - P07: 三个递进实验 — goal loop、timer loop、maker-checker loop - 配套代码模板：goal-template.md、loop-state-template.md、maker-prompt.md、checker-prompt.md - Reference 页新增 Addy Osmani Loop Engineering (2026-06-07) 到 2026 扩展参考 - README 更新徽章（13 lectures / 7 projects），学习路径图、syllabus、project 表格、changelog Build: ✓
1.6 KiB
1.6 KiB
Welcome to Projects
This is the hands-on section of Learn Harness Engineering. Reading the lectures isn't enough—you need to build the environments yourself and observe how Codex, Claude Code, or other AI agents behave under different rules.
Project Overview
This course features 7 progressive, hands-on projects that teach you how to build a reliable agentic working environment from scratch:
- Prompt-Only vs. Rules-First: Compare how an agent performs with just a prompt versus a basic harness.
- Agent-Readable Workspace: Learn how to structure your repository to make it AI-friendly and establish handoff mechanisms.
- Multi-Session Continuity: Design state files and initialization scripts so your agent can resume work seamlessly across sessions.
- Runtime Feedback and Scope Control: Introduce tools that allow the agent to test its own code and correct errors during execution.
- Self-Verification and Role Separation: Build an independent review mechanism to prevent hallucinations and early declarations of victory.
- Complete Harness (Capstone): Assemble a final, observable, end-to-end agent working environment.
- Your First Automated Loop: Transition from manual driving to automated looping — three progressive experiments: goal loop, timer loop, and maker-checker loop.
How to Proceed
Each project folder typically contains:
starter/: Your starting workspace.solution/: A reference implementation (if you get stuck).- Task instructions detailing your background and specific goals.
Use your preferred AI Coding Agent (e.g., Claude Code, Cursor, Trae) to complete the tasks inside the starter/ directory.