{
  "markdown": "<p align=\"center\">\n  <img src=\"harness_banner.png\" alt=\"Harness Banner\" width=\"600\">\n</p>\n\n<p align=\"center\">\n  <img src=\"https://img.shields.io/badge/Version-1.2.0-brightgreen.svg\" alt=\"Version\">\n  <a href=\"LICENSE\"><img src=\"https://img.shields.io/badge/License-Apache_2.0-blue.svg\" alt=\"License\"></a>\n  <img src=\"https://img.shields.io/badge/Claude_Code-Plugin-purple.svg\" alt=\"Claude Code Plugin\">\n  <img src=\"https://img.shields.io/badge/Patterns-6_Architectures-orange.svg\" alt=\"6 Architecture Patterns\">\n  <img src=\"https://img.shields.io/badge/Mode-Agent_Teams-green.svg\" alt=\"Agent Teams\">\n  <a href=\"https://github.com/revfactory/harness/stargazers\"><img src=\"https://img.shields.io/github/stars/revfactory/harness?style=social\" alt=\"GitHub Stars\"></a>\n</p>\n\n<p align=\"center\">\n  <a href=\"#category--where-harness-sits\"><img src=\"https://img.shields.io/badge/Layer-L3%20Meta--Factory-orange\" alt=\"Layer\"></a>\n  <a href=\"#category--where-harness-sits\"><img src=\"https://img.shields.io/badge/Sub--layer-Team--Architecture%20Factory-teal\" alt=\"Sub-layer\"></a>\n  <a href=\"#\"><img src=\"https://img.shields.io/badge/README-EN%20%7C%20KO%20%7C%20JA-lightgrey\" alt=\"i18n\"></a>\n</p>\n\n# Harness — The Team-Architecture Factory for Claude Code\n\n**English** | [한국어](README_KO.md) | [日本語](README_JA.md)\n\n> **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.\n\n## Overview\n\nHarness 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.\n\n## Category — Where Harness Sits\n\nHarness 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**.\n\n| Layer | What it does | Neighbors we coexist with |\n|-------|--------------|---------------------------|\n| **L3 — Meta-Factory / Team-Architecture Factory** (us) | Domain sentence → agent team + skills, via 6 pre-defined team patterns | — |\n| L3 — Meta-Factory / Runtime-Configuration Factory | Deterministic, repeatable runtime configurations | [coleam00/Archon](https://github.com/coleam00/Archon) |\n| L3 — Meta-Factory / Codex Runtime Port | Same concept, Codex runtime | [SaehwanPark/meta-harness](https://github.com/SaehwanPark/meta-harness) |\n| L2 — Cross-Harness Workflow | Standardize skills/rules/hooks across multiple harnesses | [affaan-m/ECC](https://github.com/affaan-m/everything-claude-code) |\n\n> 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.\n\n## Star History\n\n<a href=\"https://www.star-history.com/?repos=revfactory%2Fharness&type=date&legend=top-left\">\n <picture>\n   <source media=\"(prefers-color-scheme: dark)\" srcset=\"https://api.star-history.com/chart?repos=revfactory/harness&type=date&theme=dark&legend=top-left\" />\n   <source media=\"(prefers-color-scheme: light)\" srcset=\"https://api.star-history.com/chart?repos=revfactory/harness&type=date&legend=top-left\" />\n   <img alt=\"Star History Chart\" src=\"https://api.star-history.com/chart?repos=revfactory/harness&type=date&legend=top-left\" />\n </picture>\n</a>\n\n\n## Key Features\n\n- **Agent Team Design** — 6 architectural patterns: Pipeline, Fan-out/Fan-in, Expert Pool, Producer-Reviewer, Supervisor, and Hierarchical Delegation\n- **Skill Generation** — Auto-generates skills with Progressive Disclosure for efficient context management\n- **Orchestration** — Inter-agent data passing, error handling, and team coordination protocols\n- **Validation** — Trigger verification, dry-run testing, and with-skill vs without-skill comparison tests\n\n\n## Workflow\n\n```\nPhase 1: Domain Analysis\n    ↓\nPhase 2: Team Architecture Design (Agent Teams vs Subagents)\n    ↓\nPhase 3: Agent Definition Generation (.claude/agents/)\n    ↓\nPhase 4: Skill Generation (.claude/skills/)\n    ↓\nPhase 5: Integration & Orchestration\n    ↓\nPhase 6: Validation & Testing\n```\n\n## Installation\n\n### Via Marketplace\n\n#### Add the marketplace\n```shell\n/plugin marketplace add revfactory/harness\n```\n\n#### Install the plugin\n```shell\n/plugin install harness@harness-marketplace\n```\n\n### Direct Installation as Global Skill\n\n```shell\n# Copy the skills directory to ~/.claude/skills/harness/\ncp -r skills/harness ~/.claude/skills/harness\n```\n\n## Plugin Structure\n\n```\nharness/\n├── .claude-plugin/\n│   └── plugin.json                 # Plugin manifest\n├── skills/\n│   └── harness/\n│       ├── SKILL.md                # Main skill definition (6-Phase workflow)\n│       └── references/\n│           ├── agent-design-patterns.md   # 6 architectural patterns\n│           ├── orchestrator-template.md   # Team/subagent orchestrator templates\n│           ├── team-examples.md           # 5 real-world team configurations\n│           ├── skill-writing-guide.md     # Skill authoring guide\n│           ├── skill-testing-guide.md     # Testing & evaluation methodology\n│           └── qa-agent-guide.md          # QA agent integration guide\n└── README.md\n```\n\n## Usage\n\nTrigger in Claude Code with prompts like:\n\n```\nBuild a harness for this project\nDesign an agent team for this domain\nSet up a harness\n```\n\n### Execution Modes\n\n| Mode | Description | Recommended For |\n|------|-------------|-----------------|\n| **Agent Teams** (default) | TeamCreate + SendMessage + TaskCreate | 2+ agents requiring collaboration |\n| **Subagents** | Direct Agent tool invocation | One-off tasks, no inter-agent communication needed |\n\n<p align=\"center\">\n  <img src=\"harness_team.png\" alt=\"Harness Agent Team\" width=\"500\">\n</p>\n\n### Architecture Patterns\n\n| Pattern | Description |\n|---------|-------------|\n| Pipeline | Sequential dependent tasks |\n| Fan-out/Fan-in | Parallel independent tasks |\n| Expert Pool | Context-dependent selective invocation |\n| Producer-Reviewer | Generation followed by quality review |\n| Supervisor | Central agent with dynamic task distribution |\n| Hierarchical Delegation | Top-down recursive delegation |\n\n## Output\n\nFiles generated by Harness:\n\n```\nyour-project/\n├── .claude/\n│   ├── agents/          # Agent definition files\n│   │   ├── analyst.md\n│   │   ├── builder.md\n│   │   └── qa.md\n│   └── skills/          # Skill files\n│       ├── analyze/\n│       │   └── SKILL.md\n│       └── build/\n│           ├── SKILL.md\n│           └── references/\n```\n\n## Use Cases — Try These Prompts\n\nCopy any prompt below into Claude Code after installing Harness:\n\n**Deep Research**\n```\nBuild a harness for deep research. I need an agent team that can investigate\nany topic from multiple angles — web search, academic sources, community\nsentiment — then cross-validate findings and produce a comprehensive report.\n```\n\n**Website Development**\n```\nBuild a harness for full-stack website development. The team should handle\ndesign, frontend (React/Next.js), backend (API), and QA testing in a\ncoordinated pipeline from wireframe to deployment.\n```\n\n**Webtoon / Comic Production**\n```\nBuild a harness for webtoon episode production. I need agents for story\nwriting, character design prompts, panel layout planning, and dialogue\nediting. They should review each other's work for style consistency.\n```\n\n**YouTube Content Planning**\n```\nBuild a harness for YouTube content creation. The team should research\ntrending topics, write scripts, optimize titles/tags for SEO, and plan\nthumbnail concepts — all coordinated by a supervisor agent.\n```\n\n**Code Review & Refactoring**\n```\nBuild a harness for comprehensive code review. I want parallel agents\nchecking architecture, security vulnerabilities, performance bottlenecks,\nand code style — then merging all findings into a single report.\n```\n\n**Technical Documentation**\n```\nBuild a harness that generates API documentation from this codebase.\nAgents should analyze endpoints, write descriptions, generate usage\nexamples, and review for completeness.\n```\n\n**Data Pipeline Design**\n```\nBuild a harness for designing data pipelines. I need agents for schema\ndesign, ETL logic, data validation rules, and monitoring setup that\ndelegate sub-tasks hierarchically.\n```\n\n**Marketing Campaign**\n```\nBuild a harness for marketing campaign creation. The team should research\nthe target market, write ad copy, design visual concepts, and set up\nA/B test plans with iterative quality review.\n```\n\n## Coexistence — Harness and Neighbors\n\nHarness is not alone in the Claude Code / agent-framework ecosystem. The following repos live in adjacent layers; each is described in a parallel \"X is …, Harness is …\" form so you can pick the one that fits your need or combine several.\n\n| Repo | Their position | Relationship to Harness |\n|------|----------------|-------------------------|\n| [coleam00/Archon](https://github.com/coleam00/Archon) | \"harness builder\" — deterministic, repeatable runtime configurations | **Same L3, neighbor sub-layer.** Archon is a Runtime-Configuration Factory, Harness is a Team-Architecture Factory. Pick Archon for runtime determinism, Harness for team architecture, or combine them. |\n| [SaehwanPark/meta-harness](https://github.com/SaehwanPark/meta-harness) | Codex port of the same concept | **Same L3, different runtime.** Use Harness on Claude Code, meta-harness on Codex. |\n| [affaan-m/ECC](https://github.com/affaan-m/everything-claude-code) | \"Agent harness performance & workflow layer\" (sits on top of existing harnesses) | **Different layer.** ECC is a standardization layer across harnesses; Harness is a factory that generates harnesses. Serial combination possible. |\n| [wshobson/agents](https://github.com/wshobson/agents) | Subagent / skill catalog (182 agents, 149 skills) | **Factory ↔ parts supply.** wshobson is a catalog to shop from; Harness designs the team. Absorb wshobson entries as parts inside a Harness-generated team. |\n| [LangGraph](https://langchain-ai.github.io/langgraph/) | State-graph orchestration, LLM-agnostic | **Different track.** LangGraph is for long-running, state-recoverable orchestration; Harness is for fast Claude-Code-native team design. |\n\n## Built with Harness\n\n### Harness 100\n\n**[revfactory/harness-100](https://github.com/revfactory/harness-100)** — 100 production-ready agent team harnesses across 10 domains, available in both English and Korean (200 packages total). Each harness ships with 4-5 specialist agents, an orchestrator skill, and domain-specific skills — all generated by this plugin. 1,808 markdown files covering content creation, software development, data/AI, business strategy, education, legal, health, and more.\n\n### Research: A/B Testing Harness Effectiveness\n\n**[revfactory/claude-code-harness](https://github.com/revfactory/claude-code-harness)** — A controlled experiment across 15 software engineering tasks measuring the impact of structured pre-configuration on LLM code agent output quality.\n\n| Metric | Without Harness | With Harness | Improvement |\n|--------|:-:|:-:|:-:|\n| Average Quality Score | 49.5 | 79.3 | **+60%** |\n| Win Rate | — | — | **100%** (15/15) |\n| Output Variance | — | — | **-32%** |\n\nKey finding: effectiveness scales with task complexity — the harder the task, the greater the improvement (+23.8 Basic, +29.6 Advanced, +36.2 Expert).\n\n**Exact phrasing to use everywhere:** +60% avg quality (49.5 → 79.3), 15/15 win-rate, −32% variance (n=15, author-measured A/B, third-party replications pending).\n\n> Full paper: *Hwang, M. (2026). Harness: Structured Pre-Configuration for Enhancing LLM Code Agent Output Quality.*\n\n## Requirements\n\n- [Agent Teams enabled](https://code.claude.com/docs/en/agent-teams): `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1`\n\n## FAQ\n\n<details>\n<summary><b>Q1. Isn't \"+60%\" oversold?</b></summary>\n\n**A.** The +60% figure comes from an **author-measured A/B (n=15, 15 tasks, measured on the sister repo `claude-code-harness`)**. Every citation in this repo pairs the number with the disclosure \"n=15, author-measured, third-party replications pending\" in the same sentence. For adoption decisions, we recommend running a 2–4 week internal pilot and measuring your own numbers.\n\n**Evidence:**\n- Author A/B: [revfactory/claude-code-harness](https://github.com/revfactory/claude-code-harness)\n- Paper: *Hwang, M. (2026). Harness: Structured Pre-Configuration for Enhancing LLM Code Agent Output Quality*\n</details>\n\n<details>\n<summary><b>Q2. Why \"harness factory\" and not \"harness builder\"? Isn't this competing with Archon?</b></summary>\n\n**A.** Archon generates deterministic runtime configurations — it's a **Runtime-Configuration Factory**. Harness generates agent team architectures (team structure, message protocols, review gates) — it's a **Team-Architecture Factory**. They are **neighbor sub-layers of the same L3 Meta-Factory** and serve different needs. Pick Archon for runtime determinism, Harness for team-architecture patterns, or combine them (design architecture with Harness → deploy runtime with Archon).\n\n**Evidence:**\n- Archon self-definition: [clawfit docs/reference-levels.md](https://github.com/hongsw/clawfit/blob/main/docs/reference-levels.md)\n- Sub-layer declaration: see the **Category — Where Harness Sits** section above\n- Archon repo: [github.com/coleam00/Archon](https://github.com/coleam00/Archon)\n</details>\n\n<details>\n<summary><b>Q3. Isn't \"Claude Code only\" too narrow? What about Gemini/Codex?</b></summary>\n\n**A.** Currently the official runtime is Claude Code only. A Codex port of the same concept — [SaehwanPark/meta-harness](https://github.com/SaehwanPark/meta-harness) — is already public, so Codex teams can start there. Harness chose \"Claude-Code-native, deep\" over \"multi-runtime, shallow\"; cross-runtime collaboration with sibling repos (meta-harness, harness-init, OpenRig) is on the roadmap.\n\n**Evidence:**\n- Codex port: [github.com/SaehwanPark/meta-harness](https://github.com/SaehwanPark/meta-harness)\n- Cross-runtime scaffolder: [github.com/Gizele1/harness-init](https://github.com/Gizele1/harness-init)\n</details>\n\n## License\n\nApache 2.0\n",
  "bytes": 14556,
  "sha": "ab72bd9b63dd92d5e5cb375be92c668ef20cf4b2e00ad4e72eb8958706ac1eb1",
  "repo_slug": "revfactory/harness",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_revfactory_harness_harness_9d5a7755/readme"
}