Skip to main content 홈 크리에이터 akillness jeo-skills harness
harness Design domain-specific agent teams, define specialized agents, and generate the skills they use. Use when you need to decompose a complex project into coordinated multi-agent teams, choose the right architecture pattern (pipeline, fan-out/fan-in, expert pool, producer-reviewer, supervisor, hierarchical delegation), generate .claude/agents/ and .claude/skills/ files, or validate and iterate on generated harnesses. Triggers on: harness, build a harness, design agent team, agent team architecture, multi-agent skill generation, set up harness, harness engineering, domain agent team, harness for this project.
설치로 이동 Skills Marketplace 커뮤니티가 만든 AI 스킬을 발견하고 탐색하세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/akillness/jeo-skills --skill harness명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
Zip 다운로드 다운로드 중... Drive Godogen (htdt/godogen), the MIT-licensed publish-time generator that turns a game description into an autonomous Claude Code or Codex build for Godot 4 C#, Bevy Rust, or Babylon.js TypeScript. Route one request to one mode: preflight the toolchain and API keys; publish a fresh game repository or safely refresh a matching existing runtime with `./publish.sh --engine ...`; run the build and prove it from the live game or a 15-20s recording; budget paid Gemini, Grok, and Tripo3D asset generation; apply engine-specific build and capture rules; troubleshoot rendering and capture failures; or contribute through the issue-first upstream process. Use when the user wants an agent to build a playable game end to end with Godogen. Triggers on: godogen, htdt/godogen, publish.sh --engine, autonomous game development, Godot C# agent build, Bevy agent build, Babylon.js agent game, asset-gen, Tripo3D rig, proof video.
Install, route, and operate zenstory-ai/drama-skills, the MIT-licensed 10-skill creator-first suite for Chinese short dramas and motion comics. Use when the user wants to import or troubleshoot the suite; initialize or resume a filesystem project; analyze a novel; develop an adaptation; write episodes; build visual assets; produce image prompts, storyboards, or video prompts; review a project; open its local Dashboard; or run confirm-gated image, video, TTS, or music production. Route each request to the correct `short-drama-*` owner while preserving the five-document episode contract. Triggers on: drama-skills, zenstory-ai/drama-skills, short-drama, Chinese short drama, motion comic, creator-first drama workflow, 剧本, 视觉设定, 分镜, 图片提示词, 视频提示词. Route generic programmable-video work to `video-production`, webtoon panel production to `webtoon-harness`, and the OpenStory codebase to `openstory`.
Drive Mole (`mo`), tw93's GPL-3.0 macOS maintenance CLI that cleans caches and app leftovers, uninstalls apps with their remnants, purges rebuildable project artifacts, removes downloaded installers, explores disk usage, runs bounded system optimization, and reports live health. Routes one request to one mode: run a command safely (`--dry-run` first, the user runs the destructive step), consume the JSON/NDJSON agent surfaces (`mo analyze --json`, `mo status --json` / `--watch`, `mo history --json`, `~/.config/mole/clean-list.txt`), install/update/remove on the right channel, configure whitelists and scan paths, troubleshoot, or contribute to the repo. Use when a user wants to free Mac disk space or fully uninstall a Mac app. Triggers on: mole, `mo clean`, `mo uninstall`, `mo analyze`, `mo purge`, `mo status`, tw93/Mole, mole.fit, clean my Mac, what is eating my disk, CleanMyMac / AppCleaner / DaisyDisk alternative, brew install mole.
skill-autoresearch-harness name harness description Design domain-specific agent teams, define specialized agents, and generate the skills they use. Use when you need to decompose a complex project into coordinated multi-agent teams, choose the right architecture pattern (pipeline, fan-out/fan-in, expert pool, producer-reviewer, supervisor, hierarchical delegation), generate .claude/agents/ and .claude/skills/ files, or validate and iterate on generated harnesses. Triggers on: harness, build a harness, design agent team, agent team architecture, multi-agent skill generation, set up harness, harness engineering, domain agent team, harness for this project.
allowed-tools Bash Read Write Edit Glob Grep WebFetch Agent compatibility Requires Claude Code with CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1 for agent team execution. Works with Codex CLI and Gemini CLI for skill generation only. license Apache-2.0 metadata {"tags":"harness, agent-team, multi-agent, skill-generation, orchestration, meta-skill, claude-code","version":"1.0","source":"https://github.com/revfactory/harness"}
harness - Agent Team & Skill Architect
Keyword : harness · build a harness · design agent team · harness engineering
Meta-skill: harness designs the teams and skills that run your domain work.
Harness decomposes complex tasks into coordinated teams of specialized agents. It analyzes your domain, selects the right architecture pattern, generates agent definition files and skills, then validates the harness end-to-end.
Agent Teams require : CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1
When to use this skill
Decompose a complex project into a coordinated multi-agent team
Choose the right architecture: pipeline, fan-out/fan-in, expert pool, producer-reviewer, supervisor, or hierarchical delegation
Generate .claude/agents/{name}.md agent definition files
Generate .claude/skills/{name}/SKILL.md skill files with bundled resources
Validate trigger conditions, dry-run teams, and compare with/without harness quality
Build harnesses for: research, coding, content creation, code review, data pipelines, marketing
Instructions
Step 1: Domain Analysis Analyze the task and project context:
Read the codebase or user request to identify the domain, sub-tasks, and outputs
Detect user expertise level (beginner → detailed scaffolding; expert → lean definitions)
List the distinct task types — these map to agent roles
Decide execution mode:
Agent Team (default): 2+ agents need to communicate or cross-validate → use TeamCreate + SendMessageSub-agents (lightweight): tasks are independent and results only return to orchestrator → use Agent tool
Step 2: Team Architecture Design Pattern When to use Example Pipeline Sequential dependent stages Design → Code → Review → Deploy Fan-out/Fan-in Parallel independent work merged at the end 4 researchers → synthesizer Expert Pool Dynamic routing by input type Route to security, perf, or style expert Producer-Reviewer Generation + validation cycle Writer + Editor loop Supervisor Central coordinator with dynamic assignment Supervisor + migrators Hierarchical Delegation Recursive decomposition (max 2 levels) PM → Tech Lead → Engineers
Agent teams are the default ; choose sub-agents only when no inter-agent communication is needed
All agents must be file-based (.claude/agents/{name}.md) — never embed roles inline in Agent tool prompts
Avoid nesting teams (team members cannot themselves create teams)
Maximum 2 levels for hierarchical delegation
Step 3: Generate Agent Definition Files Create .claude/agents/{agent-name}.md for each agent. Use this template:
---
name: {agent-name}
description: {role and activation conditions}
model: opus
allowed-tools: {tool list}
---
# {Agent Name}
## Core Responsibilities
- {primary responsibility 1}
- {primary responsibility 2}
## Operational Principles
1. {principle 1}
2. {principle 2}
## Input Protocol
- Receives: {what inputs this agent consumes}
- Format: {expected format}
## Output Protocol
- Produces: {what outputs this agent delivers}
- Format: {output format and location}
## Error Handling
- On failure: {recovery behavior}
- Escalation: {when to notify orchestrator}
## Team Communication
- Reports to: {orchestrator or peer}
- Communicates with: {peer agents via SendMessage}
- Completion signal: {how to signal done}
Use model: opus for all agents by default unless speed is critical.
Step 4: Generate Skill Files For each skill the team needs, create .claude/skills/{skill-name}/SKILL.md following the Agent Skills spec:
Write a "pushy" description — actively invites triggering with specific conditions and synonyms
Explain why (context), not just what (commands)
Keep SKILL.md under 500 lines — move detailed docs to references/
Bundle reusable logic in scripts/
Apply progressive disclosure: metadata → body → references
Step 5: Integration & Orchestration Define the full workflow in the orchestrator skill or agent:
Before running : ensure CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1 is set in your environment for agent team mode.
Specify phase dependencies (which agents must complete before the next phase starts)
Define data passing: use absolute paths anchored to _workspace/ for intermediate artifacts
Add error handling for agent failures, timeouts, and data conflicts
Preserve intermediate artifacts for post-execution verification — do not delete them
Step 6: Validation & Testing Run validation to verify the harness before use:
bash scripts/validate-harness.sh .claude/agents/ .claude/skills/
Structure : required sections present in agent definition filesTrigger conditions : 20 eval queries (10 should-trigger + 10 should-NOT-trigger)Dry run : simulate team execution without running real tasksComparative : quality with harness vs without (baseline) — target +50% score improvement
Examples
Example 1: Research Harness (Fan-out/Fan-in) Prompt: "Build a harness for deep technology research"
Generated agents:
.claude/agents/official-researcher.md — documentation & official sources
.claude/agents/media-researcher.md — investment trends & news
.claude/agents/community-researcher.md — social response & forums
.claude/agents/background-researcher.md — competitive landscape
.claude/agents/research-orchestrator.md — synthesizes findings
Pattern: Fan-out/Fan-in (4 parallel researchers → orchestrator)
Example 2: Code Review Harness (Expert Pool) Prompt: "Design an agent team for thorough code review"
Generated agents:
.claude/agents/security-reviewer.md — OWASP, injection, secrets
.claude/agents/performance-reviewer.md — complexity, memory, latency
.claude/agents/testing-reviewer.md — coverage, assertions, mocks
.claude/agents/review-orchestrator.md — consolidates findings
Key feature: reviewers communicate directly (cross-domain issue detection)
Example 3: Content Production Harness (Pipeline + Parallel) Prompt: "Build a harness for webtoon production"
Phase 1 (parallel): worldbuilder + character-designer + plot-architect
Phase 2 (sequential): prose-stylist writes based on Phase 1
Phase 3 (parallel): science-consultant + continuity-manager review
Phase 4 (sequential): prose-stylist incorporates feedback
Best practices
File-based agents always — embedding roles inline in Agent tool prompts prevents reuse across sessionsAgent teams by default — prefer TeamCreate + SendMessage over sub-agents for any work requiring coordination"Pushy" descriptions — passive descriptions mean skills never activate; write active invitationsExplain why, not just what — agents follow reasoning better than rigid commandsPreserve intermediate artifacts — _workspace/ files enable post-run verification and debuggingValidate before deploy — run trigger eval (20 queries) and dry-run before using the harness in productionMax 2 hierarchy levels — deeper nesting creates coordination overhead without quality gains
References 