Token导航 LogoToken导航TokenDH.com
研究检索操作浏览器github未标认证来源可访问许可证需确认审计异常

omu奥姆

Agent Skill

omu 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

214

周安装

9

GitHub Stars

1

下载量

75
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:omu(奥姆)
来源仓库:https://github.com/akillness/oh-my-unity3d
仓库路径:skills/omu
安装命令:
npx skills add https://github.com/akillness/oh-my-unity3d --skill omu
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/akillness/oh-my-unity3d --skill omu

简介

omu 用于查找、检索和筛选相关信息。

  • 适合在 Codex、Claude、Cursor、Gemini CLI 中根据关键词快速定位候选结果。
  • 通过 npx skills add 命令从指定仓库安装,需确认权限和维护状态。
  • 使用前建议核实是否会触发联网、命令执行或文件读写操作。
  • 可结合原始 README 进一步验证具体用法和功能边界。

SKILL.md

OMU - Integrated Agent Orchestration

Keyword: omu | Platforms: Claude Code, Codex CLI, Gemini CLI, OpenCode Workflow: Plan (ralph + plannotator) -> Execute (team or bmad) -> Verify (unity-mcp / agent-browser) -> Cleanup (worktree-cleanup) Game Lifecycle: 기획 (Plan) → 개발 (Execute) → QA (Verify) → 수익성 (Monetize)

OMU is the release-oriented orchestration skill package shipped in this repository. It standardizes one path through planning, implementation, Unity3D verification, and worktree cleanup across the supported AI coding tools.

The .omu/ folder is the persistent game project management hub — long-term plans, short-term sprints, progress tracking, and history archiving are all managed here across OMU workflow runs.

agentation, annotate, and agentui are intentionally removed in v2.0.0. UI review is handled with agent-browser snapshots and normal edit loops only.

Quick Start

bash scripts/install.sh --all
bash scripts/check-status.sh

Then activate OMU with a task statement such as omu "ship the inventory prototype" and follow the gated PLAN -> EXECUTE -> VERIFY -> CLEANUP flow.

When to use this skill

  • When the user wants a Unity3D-oriented orchestration flow spanning planning, execution, verification, and cleanup
  • When the task should combine ralph, plannotator, team or bmad, browser verification, and optional Unity MCP verification
  • When the user invokes omu or asks for a release-oriented, cross-platform game-development workflow

Instructions

Execution contract

Run the phases in order. Do not skip PLAN. Do not enter EXECUTE without an approved plan.

Unity3D 모드 (unity-mcp 감지)

Before starting PLAN, check unity-mcp availability:

curl -sf http://localhost:8080/health >/dev/null 2>&1 && echo "unity-mcp: available" || echo "unity-mcp: not available"

If available (exit 0): activate Unity3D verification loop in VERIFY phase. If unavailable: use standard agent-browser verification.

STEP 0: Bootstrap

Create the working state directories:

mkdir -p .omc/state .omc/plans .omc/logs

STEP 0.1:.omu Game Plan Bootstrap

Before planning, check and initialize the .omu/ game management folder.

0.1a — Create .omu/ if it doesn't exist

if [ ! -d ".omu" ]; then
  mkdir -p .omu/history
  echo "(.omu folder created — populate long-term-plan.md and short-term-plan.md)"
fi

If .omu/long-term-plan.md does not exist, create it from the template at .unity-skills/omu/templates/.omu/long-term-plan.md (or use the built-in template below).

0.1b — Read existing plans

Read all .omu/ documents in parallel and build context:

FilePurposeAction
.omu/long-term-plan.md기획 컨셉, 규칙, 게임성Read → extract current concept and constraints
.omu/short-term-plan.md시스템, 밸런스, 배치, 연출Read → identify sprint scope and backlog
.omu/progress.md진행내용 체크리스트Read → find unchecked items and blockers

0.1c — Detect current game development stage

Based on .omu/progress.md content, detect and log the active stage:

기획 (Planning)  → GDD not complete, concept not finalized
개발 (Development) → Core loop implemented, systems in progress
QA (Quality Assurance) → Implementation complete, testing in progress
수익성 (Monetization) → QA passed, analytics/IAP integration in progress

Update .omc/state/omu-state.json:

{
  "game_stage": "기획 | 개발 | QA | 수익성",
  "omu_docs_loaded": true
}

User-facing message:

OMU activated. Game Stage: [stage]. Phase: PLAN. Plan docs loaded from.omu/ — [N] pending items found.

If .omc/state/omu-state.json does not exist, create it with:

{
  "mode": "omu",
  "phase": "plan",
  "task": "<detected task>",
  "plan_approved": false,
  "plan_gate_status": "pending",
  "team_available": null,
  "retry_count": 0,
  "last_error": null,
  "checkpoint": null,
  "created_at": "<ISO 8601>",
  "updated_at": "<ISO 8601>"
}

User-facing activation message:

OMU activated. Phase: PLAN.

STEP 0.1: Recovery Rules

  • Update checkpoint whenever a phase starts.
  • Update last_error and increment retry_count before failing pre-flight.
  • Resume from .omc/state/omu-state.json if a prior run already exists:

- phase=plan → re-run plannotator gate if plan_approved=false - phase=execute → skip PLAN, resume from last implementation checkpoint - phase=verify → skip PLAN+EXECUTE, re-run verification commands - phase=cleanup → run cleanup script, then mark done

  • If retry_count >= 3, stop and ask the user whether to continue.

1. PLAN

PLAN is mandatory in every OMU run.

Required output

Write plan.md with:

  • goal
  • implementation steps
  • risks
  • completion criteria

Required plan gate

Run the blocking plannotator loop:

bash scripts/plannotator-plan-loop.sh plan.md /tmp/plannotator_feedback.txt 3

Rules:

  • Never use &.
  • If plannotator is missing, auto-run bash scripts/ensure-plannotator.sh first.

Exit code handling:

  • exit 0 → approved=true → proceed to EXECUTE
  • exit 10 → approved=false (feedback) → update plan.md and re-run the loop
  • exit 32 → plannotator unavailable (non-interactive env, port conflict, or install failure) → HALT: do NOT proceed to EXECUTE. Output: "⚠️ PLAN GATE: plannotator가 필요합니다. bash scripts/ensure-plannotator.sh 실행 후 재시도하세요." TUI 폴백은 비활성화되어 있습니다. plannotator 없이 plan을 승인할 수 없습니다.

When approved, update the state file:

{
  "phase": "execute",
  "plan_approved": true,
  "plan_gate_status": "approved"
}

2. EXECUTE

OMU supports two execution paths.

Claude Code

Use team execution:

/omc:team 3:executor "<task>"

OMU expects CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1.

Codex, Gemini CLI, OpenCode

Use BMAD as the fallback orchestration path:

/workflow-init
/workflow-status

Execution rule:

  • keep the implementation aligned to the approved plan
  • update phase to execute
  • return to PLAN only if the plan is invalidated by new information

When execution is complete, update the state file:

{
  "phase": "verify",
  "checkpoint": "execute-complete"
}

.omu Progress Update (EXECUTE 단계 완료 시)

After each completed task during EXECUTE, update .omu/progress.md:

  1. Mark completed items with - [x]
  2. Add newly discovered tasks under the appropriate section
  3. Update short-term-plan.md if sprint scope changes
# Example: mark a system as complete
# .omu/progress.md: "- [ ] 코어 루프 구현" → "- [x] 코어 루프 구현"

New short-term tasks discovered during execution should be appended to .omu/short-term-plan.md under Backlog.


3. VERIFY

Use browser verification when the task includes UI or browser behavior.

Primary command:

agent-browser snapshot http://localhost:3000

Useful follow-ups:

agent-browser screenshot http://localhost:3000 -o verify.png
playwriter --help

Verification rule:

  • confirm the changed behavior, not only page load
  • capture evidence before cleanup when the task is browser-facing
  • keep the state file in phase = "verify" until verification is complete

When verification passes, update the state file:

{
  "phase": "cleanup",
  "checkpoint": "verify-complete"
}

STEP 3: VERIFY — Unity3D 검증 루프

Unity3D 모드가 활성화된 경우 브라우저 검증 대신 아래 루프를 실행합니다.

unity-mcp: run_tests → pass/fail 집계 ② unity-mcp: read_console → Error/Exception 패턴 탐지 ③ unity-mcp: editor_state → 씬 로드 상태 확인 ④ unity-mcp: find_gameobjects → 필수 오브젝트 확인

검증 결과:

  • 모두 통과 → CLEANUP
  • 실패 → Fix 루프 (code-refactoring 또는 unity-mcp: validate_script → 재검증, 최대 3회)
  • 3회 초과 → 사용자 확인 요청

omu-state.json 업데이트:

{ "unity_verify": { "tests_passed": true, "console_errors": 0, "retry_count": 0 } }

4. CLEANUP

After verification, guard against uncommitted changes:

# Guard: warn if uncommitted changes exist
if ! git diff --quiet || ! git diff --cached --quiet; then
  echo "[OMU][CLEANUP] ⚠️ Uncommitted changes detected. Commit or stash before cleanup."
  git status --short
  exit 1
fi

Then run cleanup:

bash scripts/worktree-cleanup.sh || git worktree prune

Then set:

{
  "phase": "done",
  "checkpoint": "cleanup"
}

Cleanup rule:

  • warn before cleanup if there are uncommitted changes
  • clean extra worktrees only
  • never delete unrelated user work

.omu Archive (CLEANUP 단계)

Archive completed work from .omu/ to history:

# 1. Generate archive filename
ARCHIVE_DATE=$(date +%Y%m%d)
ARCHIVE_FILE=".omu/history/${ARCHIVE_DATE}-$(echo "$TASK" | tr ' ' '-' | tr '[:upper:]' '[:lower:]' | head -c 40).md"

# 2. Extract completed items from progress.md (lines with [x])
grep -E "^\s*- \[x\]" .omu/progress.md > /tmp/completed_items.txt

# 3. Write archive entry
cat > "$ARCHIVE_FILE" << EOF
# Archive: $TASK
**Date**: $ARCHIVE_DATE
**Stage**: $(jq -r '.game_stage' .omc/state/omu-state.json)
**Sprint**: (from short-term-plan.md)

## Completed Items
$(cat /tmp/completed_items.txt)

## Notes
(Add retrospective notes here)
EOF

# 4. Remove completed items from progress.md
sed -i '' '/^\s*- \[x\]/d' .omu/progress.md

# 5. Remove completed items from short-term-plan.md
sed -i '' '/^\s*- \[x\]/d' .omu/short-term-plan.md

.omu/ post-cleanup state:

  • progress.md — only unchecked (- []) items remain
  • short-term-plan.md — backlog updated, completed sprint items removed
  • history/YYYYMMDD-<task>.md — permanent record of what was done

Examples

Unity3D 워크플로우 예제

Workflow 1: 게임 기획 → 씬 프로토타이핑

omu "씬 프로토타이핑: <게임명>"
  [PLAN]     bmad-gds-brainstorm-game → bmad-gds-gdd
  [EXECUTE]  unity-mcp: manage_scene → manage_gameobject → manage_probuilder
  [VERIFY]   unity-mcp: run_tests → read_console → editor_state → Fix 루프 (max 3)
  [CLEANUP]

Workflow 2: 스프린트 스토리 → C# 구현

omu "스토리 구현: <스토리명>"
  [PLAN]     bmad-gds-sprint-planning → bmad-gds-create-story
  [EXECUTE]  bmad-gds-dev-story → unity-mcp: create_script → validate_script → script_apply_edits
  [VERIFY]   unity-mcp: run_tests → read_console → Fix 루프 → bmad-gds-code-review
  [CLEANUP]

Workflow 3: 에셋 파이프라인 자동화

omu "에셋 파이프라인: <에셋 종류>"
  [PLAN]     file-organization
  [EXECUTE]  unity-mcp: manage_asset → manage_texture → manage_material → manage_prefabs → batch_execute
  [VERIFY]   unity-mcp: read_console → run_tests → performance-optimization → Fix 루프
  [CLEANUP]

Workflow 4: Unity UI/비주얼 개발

design-system (Unity3D Design Guide 탐색)
  → ui-component-patterns → unity-mcp: manage_ui → manage_animation → manage_vfx
  → [VERIFY] unity-mcp: run_tests → read_console

Workflow 5: 성능 최적화 & 디버깅

omu "성능 최적화: <증상>"
  [PLAN]     log-analysis (unity-mcp: read_console) → find_gameobjects → codebase-search
  [EXECUTE]  performance-optimization → unity-mcp: manage_components → batch_execute
  [VERIFY]   unity-mcp: run_tests → read_console → bmad-gds-performance-test → Fix 루프
  [CLEANUP]

6. Quick Start

Install dependencies and helpers

bash scripts/install.sh --all

Check current environment

bash scripts/check-status.sh

Configure each platform

bash scripts/setup-claude.sh
bash scripts/setup-codex.sh
bash scripts/setup-gemini.sh
bash scripts/setup-opencode.sh

7. Installed Components

ToolPurpose
omcClaude Code team orchestration
omxOpenCode orchestration
ohmgGemini multi-agent support
bmadFallback workflow orchestration
ralphspecification-first planning loop
plannotatorblocking plan review gate
agent-browserbrowser verification
playwriteroptional browser automation helper

8. Platform Notes

Claude Code

  • setup-claude.sh enables team mode and plannotator review hooks.
  • OMU does not fall back to single-agent execution in Claude Code.

Codex CLI

  • setup-codex.sh writes developer_instructions, /prompts:omu, and a PLAN_READY notify hook.

Gemini CLI

  • setup-gemini.sh installs a plannotator-oriented AfterAgent helper and appends OMU guidance to GEMINI.md.

OpenCode

  • setup-opencode.sh registers plugins and slash commands for plan, execute, verify, and cleanup.

9. State File

Path:

.omc/state/omu-state.json

Example:

{
  "mode": "omu",
  "phase": "verify",
  "task": "Ship the OMU release package",
  "plan_approved": true,
  "plan_gate_status": "approved",
  "team_available": false,
  "retry_count": 0,
  "last_error": null,
  "checkpoint": "verify",
  "created_at": "2026-03-09T00:00:00Z",
  "updated_at": "2026-03-09T00:00:00Z"
}

10. Troubleshooting

ProblemAction
plannotator unavailableRun bash scripts/ensure-plannotator.sh or bash scripts/install.sh --with-plannotator
Claude team execution unavailableRun bash scripts/setup-claude.sh, restart Claude Code, confirm CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1
Codex prompt not foundRe-run bash scripts/setup-codex.sh
Gemini hook missingRe-run bash scripts/setup-gemini.sh
Browser verification unavailableInstall agent-browser with bash scripts/install.sh --with-browser
Worktrees remain after completionRun bash scripts/worktree-cleanup.sh and then git worktree prune

11. Release Notes

v2.0.0

  • removed agentation integration from the skill contract
  • removed annotate and agentui workflow branches
  • simplified verification around agent-browser
  • aligned setup scripts and docs to the reduced release surface

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

37.48%
按下载量换算28

Claude

27.2%
按下载量换算20

Cursor

18.56%
按下载量换算14

Gemini CLI

9.43%
按下载量换算7

安全审计

Gen Agent Trust Hub

未通过

Socket

可疑

Snyk

可疑

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills