revfactory / harness
revfactory / harness
Harness — The Team-Architecture Factory for Claude Code Harness 是 Claude Code 的团队架构工厂。
Harness is a team-architecture factory for Claude Code. Say “build a harness for this project” (English) or “하네스 구성해줘” (한국어) or “ハーネスを構成して” (日本語), and the plugin turns your domain description into an agent team and the skills they use — picked from six pre-defined team-architecture patterns. Harness 是 Claude Code 的团队架构工厂。只需说出“build a harness for this project”（英语）、“하네스 구성해줘”（韩语）或“ハーネスを構成して”（日语），该插件就会根据你的领域描述，从六种预定义的团队架构模式中选择并构建出一个智能体团队及其所需技能。
Overview
概述
Harness leverages Claude Code’s agent team system to decompose complex tasks into coordinated teams of specialized agents. Say “build a harness for this project” and it automatically generates agent definitions (.claude/agents/) and skills (.claude/skills/) tailored to your domain. Harness 利用 Claude Code 的智能体团队系统，将复杂任务分解为由专业智能体组成的协作团队。只需说“build a harness for this project”，它就会自动生成针对你所在领域定制的智能体定义（.claude/agents/）和技能（.claude/skills/）。
Category — Where Harness Sits
类别 — Harness 的定位
Harness lives at the L3 Meta-Factory layer of the Claude Code ecosystem — the layer that generates other harnesses rather than being one. Inside L3, we pick a specific sub-layer: Team-Architecture Factory. Harness 位于 Claude Code 生态系统的 L3 元工厂（Meta-Factory）层——这一层负责生成其他 harness，而不是自身作为一个 harness。在 L3 内部，我们选择了特定的子层：团队架构工厂（Team-Architecture Factory）。
| Layer | What it does | Neighbors we coexist with |
|---|---|---|
| 层级 | 功能 | 共存的邻近项目 |
| L3 — Meta-Factory / Team-Architecture Factory (us) | Domain sentence → agent team + skills, via 6 pre-defined team patterns | L3 — Meta-Factory / Runtime-Configuration Factory (Deterministic, repeatable runtime configurations: coleam00/Archon) |
| L3 — 元工厂 / 团队架构工厂 (我们) | 通过 6 种预定义团队模式，将领域描述转化为智能体团队 + 技能 | L3 — 元工厂 / 运行时配置工厂 (确定性、可重复的运行时配置：coleam00/Archon) |
| L3 — Meta-Factory / Codex Runtime Port | Same concept, Codex runtime | SaehwanPark/meta-harness |
| L3 — 元工厂 / Codex 运行时端口 | 相同概念，Codex 运行时 | SaehwanPark/meta-harness |
| L2 — Cross-Harness Workflow | Standardize skills/rules/hooks across multiple harnesses | affaan-m/ECC |
| L2 — 跨 Harness 工作流 | 在多个 harness 间标准化技能/规则/钩子 | affaan-m/ECC |
Archon generates deterministic runtime configurations. Harness generates team architectures (pipeline, fan-out/fan-in, expert pool, producer-reviewer, supervisor, hierarchical delegation) plus the skills agents use. Different sub-layers of the same L3. Pick Archon for runtime determinism, Harness for team architecture, or combine them. Archon 生成确定性的运行时配置。Harness 生成团队架构（流水线、扇出/扇入、专家池、生产者-评审者、监督者、分层委派）以及智能体使用的技能。它们属于同一 L3 层的不同子层。若追求运行时确定性请选择 Archon，若追求团队架构请选择 Harness，或者将两者结合使用。
Key Features
核心功能
- Agent Team Design — 6 architectural patterns: Pipeline, Fan-out/Fan-in, Expert Pool, Producer-Reviewer, Supervisor, and Hierarchical Delegation 智能体团队设计 — 6 种架构模式：流水线、扇出/扇入、专家池、生产者-评审者、监督者和分层委派。
- Skill Generation — Auto-generates skills with Progressive Disclosure for efficient context management 技能生成 — 自动生成具备“渐进式披露”（Progressive Disclosure）功能的技能，以实现高效的上下文管理。
- Orchestration — Inter-agent data passing, error handling, and team coordination protocols 编排 — 智能体间的数据传递、错误处理和团队协作协议。
- Validation — Trigger verification, dry-run testing, and with-skill vs without-skill comparison tests 验证 — 触发验证、试运行测试以及“有技能”与“无技能”的对比测试。
Workflow
工作流
Phase 1: Domain Analysis ↓ Phase 2: Team Architecture Design (Agent Teams vs Subagents) ↓ Phase 3: Agent Definition Generation (.claude/agents/) ↓ Phase 4: Skill Generation (.claude/skills/) ↓ Phase 5: Integration & Orchestration ↓ Phase 6: Validation & Testing 第一阶段：领域分析 ↓ 第二阶段：团队架构设计（智能体团队 vs 子智能体） ↓ 第三阶段：智能体定义生成 (.claude/agents/) ↓ 第四阶段：技能生成 (.claude/skills/) ↓ 第五阶段：集成与编排 ↓ 第六阶段：验证与测试
Installation
安装
Via Marketplace
通过市场安装
/plugin marketplace add revfactory/harness
/plugin install harness@harness-marketplace
Direct Installation as Global Skill
作为全局技能直接安装
# Copy the skills directory to ~/.claude/skills/harness/
cp -r skills/harness ~/.claude/skills/harness
Usage
使用方法
Trigger in Claude Code with prompts like: 在 Claude Code 中使用如下提示词触发：
- “Build a harness for this project”
- “Design an agent team for this domain”
- “Set up a harness”
Execution Modes
执行模式
| Mode | Description | Recommended For |
|---|---|---|
| 模式 | 描述 | 推荐场景 |
| Agent Teams (default) | TeamCreate + SendMessage + TaskCreate | 2+ agents requiring collaboration |
| 智能体团队 (默认) | 团队创建 + 发送消息 + 任务创建 | 需要协作的 2 个以上智能体 |
| Subagents | Direct Agent tool invocation | One-off tasks, no inter-agent communication needed |
| 子智能体 | 直接调用智能体工具 | 一次性任务，无需智能体间通信 |
Architecture Patterns
架构模式
- Pipeline: Sequential dependent tasks (流水线：顺序依赖任务)
- Fan-out/Fan-in: Parallel independent tasks (扇出/扇入：并行独立任务)
- Expert Pool: Context-dependent selective invocation (专家池：上下文相关的选择性调用)
- Producer-Reviewer: Generation followed by quality review (生产者-评审者：生成后进行质量评审)
- Supervisor: Central agent with dynamic task distribution (监督者：具备动态任务分配能力的中心智能体)
- Hierarchical Delegation: Top-down recursive delegation (分层委派：自上而下的递归委派)