Back to Details

harness-engineering

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Harness Engineering Framework

Harness Engineering框架

One-click initialization of a complete Harness Engineering framework in any project directory.
Based on insights from OpenAI (Codex), Anthropic (3-agent GAN architecture), and LangChain (self-verify loops), this skill sets up:
  • 3-agent architecture: Planner (spec), Generator (build), Evaluator (test)
  • Sprint contracts: Machine-verifiable "done" criteria before coding
  • Quality hooks: Loop detection, pre-completion checklist, context injection
  • Slash commands: 
    /plan
    , 
    /build
    , 
    /qa
    , 
    /sprint
  • Golden principles: 10 non-negotiable rules enforced across all agents
一键在任意项目目录中初始化完整的Harness Engineering框架。
该技能基于OpenAI（Codex）、Anthropic（3-agent GAN架构）和LangChain（自验证循环）的技术洞察搭建，可配置：
  • 3-Agent架构：Planner（规格制定）、Generator（构建）、Evaluator（测试）
  • Sprint契约：编码前即可机器验证的“完成”标准
  • 质量钩子：循环检测、完成前检查清单、上下文注入
  • 斜杠命令：
    /plan
    、
    /build
    、
    /qa
    、
    /sprint
  • 黄金原则：所有Agent需遵循的10条不可协商规则

When to Use

使用场景

  • Starting a new project and want structured AI-assisted development
  • Want to set up Plan-Build-Verify-Fix workflow in current project
  • User says "harness", "init harness", "setup framework", or similar
  • 启动新项目，希望采用结构化AI辅助开发流程
  • 想要在当前项目中搭建Plan-Build-Verify-Fix工作流
  • 用户提及"harness"、"init harness"、"setup framework"或类似表述

Initialization Process

初始化流程

Step 1: Gather Project Info

步骤1：收集项目信息

Before generating files, ask the user:
  1. Project name (or detect from current directory name)
  2. Tech stack (optional, e.g. "React + Node.js", "Python FastAPI", "Go microservice")
  3. Project type (web app, API service, CLI tool, library, etc.)
If the user provides a description with 
/harness <description>
, extract the info from context.
生成文件前，需询问用户：
  1. 项目名称（或从当前目录名称自动检测）
  2. 技术栈（可选，例如"React + Node.js"、"Python FastAPI"、"Go微服务"）
  3. 项目类型（Web应用、API服务、CLI工具、库等）
若用户提供
/harness <描述>
形式的内容，则从上下文提取相关信息。

Step 2: Generate Framework Files

步骤2：生成框架文件

Execute the scaffold script:
bash
python3 {{SKILL_PATH}}/scripts/scaffold.py --project-name "<PROJECT_NAME>" --tech-stack "<TECH_STACK>" --project-type "<PROJECT_TYPE>" --target-dir "<CURRENT_PROJECT_DIR>"
This generates the following structure in the current project:
<project>/
  CLAUDE.md                        # Project map (<80 lines)
  .claude/
    agents/
      planner.md                   # Spec creation agent
      generator.md                 # Implementation agent
      evaluator.md                 # Testing/grading agent
      doc-gardener.md              # Doc freshness agent
    commands/
      plan.md                      # /plan command
      build.md                     # /build command
      qa.md                        # /qa command
      sprint.md                    # /sprint command
    hooks/
      loop-detector.py             # File edit loop detection
      pre-completion-check.py      # Task completion checklist
      context-injector.py          # Session context middleware
  docs/
    architecture.md                # System design
    golden-principles.md           # Non-negotiable rules
    sprint-workflow.md             # Sprint process
    contracts/
      TEMPLATE.md                  # Sprint contract template
    specs/                         # (populated by planner)
    plans/                         # (populated by planner)
执行脚手架脚本：
bash
python3 {{SKILL_PATH}}/scripts/scaffold.py --project-name "<PROJECT_NAME>" --tech-stack "<TECH_STACK>" --project-type "<PROJECT_TYPE>" --target-dir "<CURRENT_PROJECT_DIR>"
此命令会在当前项目中生成以下结构：
<project>/
  CLAUDE.md                        # 项目地图（少于80行）
  .claude/
    agents/
      planner.md                   # 规格创建Agent
      generator.md                 # 实现Agent
      evaluator.md                 # 测试/评估Agent
      doc-gardener.md              # 文档更新Agent
    commands/
      plan.md                      # /plan命令
      build.md                     # /build命令
      qa.md                        # /qa命令
      sprint.md                    # /sprint命令
    hooks/
      loop-detector.py             # 文件编辑循环检测
      pre-completion-check.py      # 任务完成检查清单
      context-injector.py          # 会话上下文中间件
  docs/
    architecture.md                # 系统设计
    golden-principles.md           # 不可协商规则
    sprint-workflow.md             # Sprint流程
    contracts/
      TEMPLATE.md                  # Sprint契约模板
    specs/                         # （由Planner填充）
    plans/                         # （由Planner填充）

Step 3: Configure Hooks

步骤3：配置钩子

After generating files, merge hook configuration into the project's 
.claude/settings.json
(or create it):
bash
python3 {{SKILL_PATH}}/scripts/merge_settings.py --target-dir "<CURRENT_PROJECT_DIR>"
This adds hook definitions without overwriting existing settings.
生成文件后，将钩子配置合并到项目的
.claude/settings.json
中（若不存在则创建）：
bash
python3 {{SKILL_PATH}}/scripts/merge_settings.py --target-dir "<CURRENT_PROJECT_DIR>"
此操作会添加钩子定义，且不会覆盖现有配置。

Step 4: Verify Installation

步骤4：验证安装

Confirm all files were created:
bash
ls -la CLAUDE.md .claude/agents/ .claude/commands/ .claude/hooks/ docs/
Report to the user what was created and how to start using it.
确认所有文件已创建：
bash
ls -la CLAUDE.md .claude/agents/ .claude/commands/ .claude/hooks/ docs/
向用户报告已创建的内容及使用方法。

Usage After Initialization

初始化后的使用方法

CommandPurpose
/plan <description>
Create a feature specification from 1-4 sentences
/build
Build the most recent spec using sprint workflow
/qa
Run evaluator against current code
/sprint <description>
Full Plan-Build-Verify cycle from scratch
命令用途
/plan <描述>
根据1-4句话的描述创建功能规格
/build
使用Sprint工作流构建最新规格
/qa
针对当前代码运行Evaluator测试
/sprint <描述>
从头开始执行完整的Plan-Build-Verify周期

Key Principles

核心原则

The framework enforces these rules (see 
docs/golden-principles.md
after install):
  1. Spec before code - No implementation without a written spec
  2. Testable criteria - Every feature has machine-verifiable acceptance criteria
  3. Self-verify first - Generator must self-check before evaluator runs
  4. Loop awareness - Editing same file 5+ times triggers a stop-and-reassess
  5. Contract-driven - Sprint contracts define "done" before coding begins
该框架强制执行以下规则（安装后可查看
docs/golden-principles.md
）：
  1. 先规格后编码 - 无书面规格则不进行实现
  2. 可测试标准 - 每个功能都有可机器验证的验收标准
  3. 先自我验证 - Generator必须先自行检查，再由Evaluator运行测试
  4. 循环感知 - 同一文件编辑5次以上时，触发停止并重新评估
  5. 契约驱动 - Sprint契约在编码前定义“完成”标准

Architecture

架构

For full details, read the generated 
docs/architecture.md
after installation. Key flow:
User Prompt → Planner (spec) → Generator + Evaluator negotiate contract
→ Generator builds → Evaluator tests → Fix loop (max 3x) → Complete
详细内容请查看安装后生成的
docs/architecture.md
。核心流程：
用户提示 → Planner（制定规格）→ Generator + Evaluator协商契约
→ Generator构建 → Evaluator测试 → 修复循环（最多3次）→ 完成