moai-adk 0.9.0__py3-none-any.whl → 0.15.0__py3-none-any.whl

moai_adk/templates/.claude/commands/alfred/1-plan.md

@@ -1,10 +1,6 @@
 ---
 name: alfred:1-plan
-description: "Planning (brainstorming, plan writing, design discussion) + Branch/PR creation"
-# Translations:
-# - ko: "계획 수립 (브레인스토밍, 설계 논의) + 브랜치/PR 생성"
-# - ja: "計画策定（ブレインストーミング、設計議論）+ ブランチ/PR作成"
-# - zh: "规划（头脑风暴、设计讨论）+ 分支/PR创建"
+description: "Define specifications and create development branch"
 argument-hint: Title 1 Title 2 ... | SPEC-ID modifications
 allowed-tools:
 - Read
@@ -22,6 +18,12 @@ allowed-tools:
 
 # 🏗️ MoAI-ADK Step 1: Establish a plan (Plan) - Always make a plan first and then proceed.
 > **Note**: Interactive prompts use `AskUserQuestion tool (documented in moai-alfred-interactive-questions skill)` for TUI selection menus. The skill is loaded on-demand when user interaction is required.
+>
+> **Batched Design**: All AskUserQuestion calls follow batched design principles (1-4 questions per call) to minimize user interaction turns. See CLAUDE.md section "Alfred Command Completion Pattern" for details.
+
+<!-- @CODE:ALF-WORKFLOW-001:CMD-PLAN -->
+
+**4-Step Workflow Integration**: This command implements Steps 1-2 of Alfred's workflow (Intent Understanding → Plan Creation). See CLAUDE.md for full workflow details.
 
 ## 🎯 Command Purpose
 
@@ -167,11 +169,11 @@ STEP 1 consists of **two independent phases** to provide flexible workflow based
 Invoking the Task tool (Explore agent):
 - subagent_type: "Explore"
 - description: "Explore related files in the codebase"
-- prompt: "Please find all files related to the following keywords: $ARGUMENTS
- - File location (src/, tests/, docs/)
- - Relevant SPEC document (.moai/specs/)
- - Existing implementation code
- thoroughness level: medium"
+- prompt: "다음 키워드와 관련된 모든 파일을 찾아주세요: $ARGUMENTS
+ - 파일 위치 (src/, tests/, docs/)
+ - 관련 SPEC 문서 (.moai/specs/)
+ - 기존 구현 코드
+ 상세도 수준: medium"
 ```
 
 **Note**: If user provides clear SPEC title, skip Phase A and proceed directly to Phase B.
@@ -190,34 +192,47 @@ This phase is **always required** regardless of whether Phase A was executed.
 Call the Task tool:
 - subagent_type: "spec-builder"
 - description: "Analyze the plan and establish a plan"
-- prompt: """You are spec-builder agent.
-
-LANGUAGE CONFIGURATION:
-- conversation_language: {{CONVERSATION_LANGUAGE}}
-- language_name: {{CONVERSATION_LANGUAGE_NAME}}
-
-CRITICAL INSTRUCTION:
-All SPEC documents and analysis must be generated in conversation_language.
-- If conversation_language is 'ko' (Korean): Generate ALL analysis, plans, and SPEC documents in Korean
-- If conversation_language is 'ja' (Japanese): Generate ALL analysis, plans, and SPEC documents in Japanese
-- If conversation_language is other language: Follow the specified language
-
-SKILL INVOCATION:
-Use explicit Skill() calls when needed:
-- Skill("moai-foundation-specs") for SPEC structure guidance
-- Skill("moai-foundation-ears") for EARS syntax requirements
-- Skill("moai-alfred-spec-metadata-validation") for metadata validation
-
-TASK:
-Please analyze the project document and suggest SPEC candidates.
-Run in analysis mode, and must include the following:
-1. In-depth analysis of product/structure/tech.md
-2. Identify SPEC candidates and Determine priorities
-3. Design EARS structure
-4. Wait for user approval
-
-User input: $ARGUMENTS
-(Optional) Explore results: $EXPLORE_RESULTS"""
+- prompt: """당신은 spec-builder 에이전트입니다.
+
+언어 설정:
+- 대화_언어: {{CONVERSATION_LANGUAGE}}
+- 언어명: {{CONVERSATION_LANGUAGE_NAME}}
+
+중요 지시사항:
+SPEC 문서는 이중 언어 구조를 따라야 합니다 (사용자 언어 + 영어 요약):
+
+conversation_language == 'ko' (한국어)인 경우:
+- YAML 메타데이터: 영어만 사용
+- 제목 (@SPEC 태그): 한국어 주요, 영어 버전은 하단에 기재
+- 주요 내용 (분석, 요구사항, EARS): 한국어
+- SUMMARY 섹션: 영어 (국제 기여자를 위해 100-200단어)
+- HISTORY: 한국어 (새로운 항목), 주요 버전에는 영어 요약
+
+conversation_language == 'ja' (일본어)인 경우:
+- 한국어와 동일한 이중 언어 패턴 사용
+- 주요 내용: 일본어
+- SUMMARY: 영어
+
+다른 언어인 경우:
+- 주요 내용: 사용자 지정 언어
+- SUMMARY: 영어 (항상)
+
+스킬 호출:
+필요 시 명시적 Skill() 호출 사용:
+- Skill("moai-foundation-specs") - SPEC 구조 가이드
+- Skill("moai-foundation-ears") - EARS 문법 요구사항
+- Skill("moai-alfred-spec-metadata-validation") - 메타데이터 검증
+
+작업:
+프로젝트 문서를 분석하여 SPEC 후보자를 제시해주세요.
+분석 모드로 실행하며, 다음을 포함해야 합니다:
+1. product/structure/tech.md의 심층 분석
+2. SPEC 후보자 식별 및 우선순위 결정
+3. EARS 구조 설계
+4. 사용자 승인 대기
+
+사용자 입력: $ARGUMENTS
+(선택사항) 탐색 결과: $EXPLORE_RESULTS"""
 ```
 
 **Note**: If Phase A was executed, pass the exploration results via `$EXPLORE_RESULTS` variable.
@@ -259,36 +274,58 @@ After user approval (collected via `AskUserQuestion tool (documented in moai-alf
 1. Call spec-builder (create plan):
    - subagent_type: "spec-builder"
    - description: "Create SPEC document"
-   - prompt: """You are spec-builder agent.
+   - prompt: """당신은 spec-builder 에이전트입니다.
 
-LANGUAGE CONFIGURATION:
-- conversation_language: {{CONVERSATION_LANGUAGE}}
-- language_name: {{CONVERSATION_LANGUAGE_NAME}}
+언어 설정:
+- 대화_언어: {{CONVERSATION_LANGUAGE}}
+- 언어명: {{CONVERSATION_LANGUAGE_NAME}}
 
-CRITICAL INSTRUCTION:
-ALL SPEC documents MUST be generated in conversation_language:
-- spec.md: Full document in conversation_language
-- plan.md: Full document in conversation_language
-- acceptance.md: Full document in conversation_language
+중요 지시사항:
+모든 SPEC 문서는 대화_언어로 작성되어야 합니다:
+- spec.md: 전체 문서를 대화_언어로 작성
+- plan.md: 전체 문서를 대화_언어로 작성
+- acceptance.md: 전체 문서를 대화_언어로 작성
 
-YAML frontmatter and @TAG identifiers MUST remain in English.
-Code examples and technical keywords can be mixed (code in English, narrative in user language).
+YAML 프론트매터와 @TAG 식별자는 반드시 영어로 유지합니다.
+코드 예제와 기술 키워드는 혼합 가능 (코드는 영어, 설명은 사용자 언어).
 
-SKILL INVOCATION:
-Use explicit Skill() calls when needed:
-- Skill("moai-foundation-specs") for SPEC structure guidance
-- Skill("moai-foundation-ears") for EARS syntax requirements
-- Skill("moai-alfred-spec-metadata-validation") for metadata validation
-- Skill("moai-alfred-tag-scanning") for TAG chain references
+스킬 호출:
+필요 시 명시적 Skill() 호출 사용:
+- Skill("moai-foundation-specs") - SPEC 구조 가이드
+- Skill("moai-foundation-ears") - EARS 문법 요구사항
+- Skill("moai-alfred-spec-metadata-validation") - 메타데이터 검증
+- Skill("moai-alfred-tag-scanning") - TAG 체인 참조
 
-TASK:
-Please fill out the SPEC document according to the plan approved in STEP 1.
-Create a specification for the EARS structure."""
+작업:
+STEP 1에서 승인된 계획에 따라 SPEC 문서를 작성해주세요.
+EARS 구조에 대한 명세를 작성합니다."""
 
 2. Invoke git-manager (Git task):
    - subagent_type: "git-manager"
-   - description: "Create Git branch/PR"
-   - prompt: "After completing the plan, please create a branch and Draft PR."
+   - description: "Create Git branch/PR with duplicate prevention"
+   - prompt: """당신은 git-manager 에이전트입니다.
+
+언어 설정:
+- 대화_언어: {{CONVERSATION_LANGUAGE}}
+- 언어명: {{CONVERSATION_LANGUAGE_NAME}}
+
+중요 지시사항 (팀 모드 중복 방지):
+GitHub Issue 또는 PR을 만들기 전에:
+1. 항상 제목에 SPEC-ID가 있는 기존 Issue를 확인하세요
+2. 항상 feature/SPEC-{ID} 브랜치명의 기존 PR을 확인하세요
+3. Issue가 존재하면 → 업데이트, 중복 생성 금지
+4. PR이 존재하면 → 업데이트, 중복 생성 금지
+5. 둘 다 존재하면 → 최신 SPEC 버전으로 모두 업데이트
+6. 레이블 필터 실패 시 대체 검색 사용 (일부 Issue는 레이블 없을 수 있음)
+7. 항상 레이블 추가: "spec", "planning", + 우선순위 레이블
+
+git-manager.md의 "SPEC 작성 시" 섹션에서 자세한 중복 방지 프로토콜과 코드 예제를 참고하세요.
+
+작업:
+완성된 SPEC 문서에 대해 기능 브랜치(feature/SPEC-{SPEC_ID})와 Draft PR(→ develop)을 생성합니다.
+GitHub 엔티티를 생성하기 전에 중복 방지 프로토콜을 구현합니다.
+
+출력 언어: {{CONVERSATION_LANGUAGE}}"""
 ```
 
 ## function
@@ -443,7 +480,7 @@ Only if the user selects **"Proceed"** or **"Start"** will Alfred call the spec-
 
 #### YAML Front Matter Schema
 
-> **📋 SPEC Metadata Standard (SSOT)**: `.moai/memory/spec-metadata.md`
+> **📋 SPEC Metadata Standard (SSOT)**: Skill("moai-alfred-spec-metadata-extended")
 
 **Metadata that must be included** at the top of the spec.md file:
 - **7 required fields**: id, version, status, created, updated, author, priority
@@ -470,7 +507,7 @@ priority: high
 - **author**: GitHub @ prefix is required before ID (e.g. `@Goos`)
 - **priority**: critical | high | medium | low
 
-**Full field description and validation methods**: see `.moai/memory/spec-metadata.md`
+**Full field description and validation methods**: see Skill("moai-alfred-spec-metadata-extended")
 
 #### HISTORY section (required)
 
@@ -504,10 +541,10 @@ You must include a HISTORY section **right after the YAML Front Matter**:
 
 **HISTORY writing rules**:
 - **Version system**: v0.0.1 (INITIAL) → v0.1.0 (implementation complete) → v1.0.0 (stabilization)
- - Detailed version system: See `.moai/memory/spec-metadata.md#version-system`
+ - Detailed version system: See Skill("moai-alfred-spec-metadata-extended") for version-system guide
 - **Version order**: Latest version on top (reverse order)
 - **Change type tag**: INITIAL, ADDED, CHANGED, IMPLEMENTATION COMPLETED, BREAKING, DEPRECATED, REMOVED, FIXED
- - Detailed description: See `.moai/memory/spec-metadata.md#history-writing-guide`
+ - Detailed description: See Skill("moai-alfred-spec-metadata-extended") for history-writing-guide
 - **Required items**: Version, date, AUTHOR, changes
 - **Optional items**: REVIEW, SCOPE, CONTEXT, MIGRATION
 
@@ -710,7 +747,7 @@ See `.coderabbit.yaml` for detailed SPEC review checklist.
 
 ## 🧠 Context Management
 
-> For more information: `.moai/memory/development-guide.md` - see section "Context Engineering"
+> For more information: Skill("moai-alfred-dev-guide") - see section "Context Engineering"
 
 ### Core strategy of this command
 
@@ -720,6 +757,43 @@ See `.coderabbit.yaml` for detailed SPEC review checklist.
 
 ---
 
+## Final Step
+
+After plan creation completes, Alfred automatically invokes AskUserQuestion to ask the user what to do next:
+
+```python
+AskUserQuestion(
+    questions=[
+        {
+            "question": "스펙 작성이 완료되었습니다. 다음으로 뭘 하시겠습니까?",
+            "header": "다음 단계",
+            "multiSelect": false,
+            "options": [
+                {
+                    "label": "🔨 구현 시작",
+                    "description": "/alfred:2-run SPEC-XXX로 TDD 구현 진행"
+                },
+                {
+                    "label": "📝 스펙 수정",
+                    "description": "SPEC 문서 검토 후 수정"
+                },
+                {
+                    "label": "🔄 새 세션 시작",
+                    "description": "성능 최적화를 위해 /clear 실행"
+                }
+            ]
+        }
+    ]
+)
+```
+
+**User Responses**:
+- **🔨 구현 시작**: Proceed to `/alfred:2-run SPEC-XXX` for TDD implementation
+- **📝 스펙 수정**: Review and modify SPEC documents before implementation
+- **🔄 새 세션 시작**: Execute `/clear` to start fresh session (recommended for performance)
+
+---
+
 ## Next steps
 
 **Recommendation**: For better performance and context management, start a new chat session with the `/clear` or `/new` command before proceeding to the next step.

moai_adk/templates/.claude/commands/alfred/2-run.md

@@ -1,10 +1,6 @@
 ---
 name: alfred:2-run
-description: "Execute planned work (TDD implementation, prototyping, documentation, etc.)"
-# Translations:
-# - ko: "계획된 작업 실행 (TDD 구현, 프로토타이핑, 문서화 등)"
-# - ja: "計画されたタスクの実行（TDD実装、プロトタイピング、ドキュメント作成など）"
-# - zh: "执行计划任务（TDD实现、原型开发、文档编写等）"
+description: "Execute TDD implementation cycle"
 argument-hint: "SPEC-ID - All with SPEC ID to implement (e.g. SPEC-001) or all \"SPEC Implementation\""
 allowed-tools:
   - Read
@@ -25,6 +21,12 @@ allowed-tools:
 
 # ⚒️ MoAI-ADK Phase 2: Run the plan - Flexible implementation strategy
 > **Note**: Interactive prompts use `AskUserQuestion tool (documented in moai-alfred-interactive-questions skill)` for TUI selection menus. The skill is loaded on-demand when user interaction is required.
+>
+> **Batched Design**: All AskUserQuestion calls follow batched design principles (1-4 questions per call) to minimize user interaction turns. See CLAUDE.md section "Alfred Command Completion Pattern" for details.
+
+<!-- @CODE:ALF-WORKFLOW-002:CMD-RUN -->
+
+**4-Step Workflow Integration**: This command implements Step 3 of Alfred's workflow (Task Execution with TodoWrite tracking). See CLAUDE.md for full workflow details.
 
 ## 🎯 Command Purpose
 
@@ -154,12 +156,12 @@ STEP 1 consists of **two independent phases** to provide flexible workflow based
 Invoking the Task tool (Explore agent):
 - subagent_type: "Explore"
 - description: "Explore existing code structures and patterns"
-- prompt: "Please explore existing code related to SPEC-$ARGUMENTS:
- - Similar function implementation code (src/)
- - Test patterns for reference (tests/)
- - Architectural patterns and design patterns
- - Current libraries and versions (package.json, requirements.txt)
- thoroughness level: medium"
+- prompt: "SPEC-$ARGUMENTS와 관련된 기존 코드를 탐색해주세요:
+ - 유사한 기능 구현 코드 (src/)
+ - 참고할 테스트 패턴 (tests/)
+ - 아키텍처 패턴 및 디자인 패턴
+ - 현재 라이브러리 및 버전 (package.json, requirements.txt)
+ 상세도 수준: medium"
 ```
 
 **Note**: If you skip Phase A, proceed directly to Phase B.
@@ -178,15 +180,15 @@ This phase is **always required** regardless of whether Phase A was executed.
 Task tool call:
 - subagent_type: "implementation-planner"
 - description: "SPEC analysis and establishment of execution strategy"
-- prompt: "Please analyze the SPEC of $ARGUMENTS and establish an execution plan.
- It must include the following:
- 1. SPEC requirements extraction and complexity assessment
- 2. Library and tool selection (using WebFetch)
- 3. TAG chain design
- 4. Step-by-step execution plan
- 5. Risks and response plans
- 6. Create action plan and use `AskUserQuestion tool (documented in moai-alfred-interactive-questions skill)` to confirm the next action with the user
- (Optional) Explore results: $EXPLORE_RESULTS"
+- prompt: "$ARGUMENTS의 SPEC을 분석하고 실행 계획을 수립해주세요.
+ 다음을 포함해야 합니다:
+ 1. SPEC 요구사항 추출 및 복잡도 평가
+ 2. 라이브러리 및 도구 선택 (WebFetch 사용)
+ 3. TAG 체인 설계
+ 4. 단계별 실행 계획
+ 5. 위험 요소 및 대응 계획
+ 6. 행동 계획을 작성하고 `AskUserQuestion 도구 (moai-alfred-interactive-questions 스킬 참고)`로 사용자와 다음 단계를 확인합니다
+ (선택사항) 탐색 결과: $EXPLORE_RESULTS"
 ```
 
 **Note**: If Phase A was executed, pass the exploration results via `$EXPLORE_RESULTS` variable.
@@ -233,6 +235,125 @@ After reviewing the action plan, select one of the following:
 
 After user approval (gathered through `AskUserQuestion tool (documented in moai-alfred-interactive-questions skill)`), **call the tdd-implementer agent using the Task tool**.
 
+---
+
+### 2.0.5 Domain Readiness Check (Automatic - Before Implementation)
+
+**Purpose**: Load domain-expert agents as "implementation advisors" based on SPEC.stack metadata
+
+**When to run**: After user approval, BEFORE invoking tdd-implementer
+
+**Detection Logic**:
+
+Alfred reads the SPEC metadata to identify required domains:
+
+```bash
+# Read SPEC metadata
+spec_metadata=$(grep "^domains:" .moai/specs/SPEC-{ID}/spec.md)
+
+# Or read from config.json
+selected_domains=$(jq -r '.stack.selected_domains[]' .moai/config.json)
+```
+
+**Domain Expert Advisory Pattern**:
+
+| Domain | Readiness Check | Advisory Focus |
+|--------|----------------|----------------|
+| **Frontend** | Component structure, testing strategy, state management | Component hierarchy, React/Vue best practices, UI testing patterns |
+| **Backend** | API contract, database schema, async patterns | RESTful design, database indexing, error handling, authentication |
+| **DevOps** | Docker readiness, environment variables, health checks | Containerization, CI/CD integration, deployment strategies |
+| **Database** | Schema design, migration strategy, indexing | Data modeling, query optimization, migration safety |
+| **Data Science** | Data pipeline design, notebook structure | ETL patterns, data validation, model versioning |
+| **Mobile** | Platform-specific requirements, app lifecycle | Native integration, state management, offline support |
+
+**Example Invocation** (Frontend + Backend detected):
+
+```python
+# Read SPEC metadata
+spec_domains = ["frontend", "backend"]  # from SPEC frontmatter
+
+# Invoke domain experts BEFORE tdd-implementer
+for domain in spec_domains:
+    if domain == "frontend":
+        Task(
+            subagent_type="Explore",
+            prompt="""You are consulting as frontend-expert for TDD implementation.
+
+SPEC: [SPEC-UI-001 - User Dashboard Component]
+
+Provide implementation readiness check:
+1. Component structure recommendations
+2. State management approach (Redux/Zustand/Context)
+3. Testing strategy (Jest + Testing Library)
+4. Accessibility requirements
+5. Performance optimization tips
+
+Output: Brief advisory for tdd-implementer (3-4 key points)"""
+        )
+
+    if domain == "backend":
+        Task(
+            subagent_type="Explore",
+            prompt="""You are consulting as backend-expert for TDD implementation.
+
+SPEC: [SPEC-API-001 - Authentication Endpoints]
+
+Provide implementation readiness check:
+1. API contract validation
+2. Database schema requirements
+3. Authentication/authorization patterns
+4. Error handling strategy
+5. Async processing considerations
+
+Output: Brief advisory for tdd-implementer (3-4 key points)"""
+        )
+```
+
+**Output Format** (Stored in SPEC plan.md):
+
+```markdown
+## Domain Expert Advisory (Implementation Phase)
+
+### Frontend Readiness
+- Component structure: Use compound component pattern for Dashboard
+- State management: Recommend Zustand for lightweight state
+- Testing: Prioritize user interaction tests over implementation details
+- Performance: Implement React.memo for expensive components
+
+### Backend Readiness
+- API contract: OpenAPI 3.0 spec generated from FastAPI
+- Database schema: Add index on user_id and created_at columns
+- Authentication: Use JWT with refresh token rotation
+- Async: Use background tasks for email notifications
+```
+
+**Integration with tdd-implementer**:
+
+```python
+# Pass domain expert feedback to tdd-implementer
+Task(
+    subagent_type="tdd-implementer",
+    prompt="""You are tdd-implementer agent.
+
+SPEC: SPEC-{ID}
+
+DOMAIN EXPERT ADVISORY:
+{domain_expert_feedback}
+
+Execute TDD implementation considering domain expert guidance.
+Follow RED → GREEN → REFACTOR cycle with domain best practices.
+
+$ARGUMENTS"""
+)
+```
+
+**Graceful Degradation**:
+- If SPEC.stack.domains missing → Skip advisory (greenfield implementation)
+- If domain expert unavailable → Continue with tdd-implementer only
+- Advisory is non-blocking (implementation proceeds regardless)
+
+---
+
 ### ⚙️ How to call an agent
 
 **STEP 2 calls tdd-implementer using the Task tool**:
@@ -241,35 +362,39 @@ After user approval (gathered through `AskUserQuestion tool (documented in moai-
 Call the Task tool:
 - subagent_type: "tdd-implementer"
 - description: "Execute task with TDD implementation"
-- prompt: """You are tdd-implementer agent.
-
-LANGUAGE CONFIGURATION:
-- conversation_language: {{CONVERSATION_LANGUAGE}}
-- language_name: {{CONVERSATION_LANGUAGE_NAME}}
-
-CRITICAL INSTRUCTION:
-Code and technical output MUST be in English.
-Code comments MAY be in {{CONVERSATION_LANGUAGE}} if appropriate.
-Test descriptions and documentation can use {{CONVERSATION_LANGUAGE}}.
-
-SKILL INVOCATION:
-Use explicit Skill() calls when needed:
-- Skill("moai-alfred-language-detection") for project language detection
-- Skill("moai-lang-python") or language-specific Skills for best practices
-- Skill("moai-essentials-debug") when tests fail
-- Skill("moai-essentials-refactor") during REFACTOR phase
-
-TASK: Execute the task according to the plan approved in STEP 1.
-
-For TDD scenario:
-- Perform RED → GREEN → REFACTOR cycle
-- Perform the following for each TAG:
-  1. RED Phase: Write a test that fails with the @TEST:ID tag
-  2. GREEN Phase: Minimal implementation with the @CODE:ID tag
-  3. REFACTOR Phase: Improve code quality
-  4. Verify TAG completion conditions and proceed to the next TAG
-
-Execute on: $ARGUMENTS"""
+- prompt: """당신은 tdd-implementer 에이전트입니다.
+
+언어 설정:
+- 대화_언어: {{CONVERSATION_LANGUAGE}}
+- 언어명: {{CONVERSATION_LANGUAGE_NAME}}
+
+중요 지시사항:
+**Code and technical output MUST be in English.** This ensures global compatibility and maintainability.
+
+코드 문법 및 키워드: 영어 (고정).
+코드 주석:
+- 로컬 프로젝트 코드: 반드시 {{CONVERSATION_LANGUAGE}}로 작성
+- 패키지 코드 (src/moai_adk/): 반드시 영어로 작성 (글로벌 배포용)
+테스트 설명 및 문서: 반드시 {{CONVERSATION_LANGUAGE}}로 작성.
+
+스킬 호출:
+필요 시 명시적 Skill() 호출 사용:
+- Skill("moai-alfred-language-detection") - 프로젝트 언어 감지
+- Skill("moai-lang-python") 또는 언어별 스킬 - 베스트 프랙티스
+- Skill("moai-essentials-debug") - 테스트 실패 시
+- Skill("moai-essentials-refactor") - REFACTOR 단계에서
+
+작업: STEP 1에서 승인된 계획에 따라 작업을 실행합니다.
+
+TDD 시나리오의 경우:
+- RED → GREEN → REFACTOR 사이클 수행
+- 각 TAG에 대해 다음을 수행:
+  1. RED 단계: @TEST:ID 태그로 실패하는 테스트 작성
+  2. GREEN 단계: @CODE:ID 태그로 최소한의 구현
+  3. REFACTOR 단계: 코드 품질 개선
+  4. TAG 완료 조건 확인 및 다음 TAG 진행
+
+실행 대상: $ARGUMENTS"""
 ```
 
 ## 🔗 TDD optimization for each language
@@ -526,7 +651,7 @@ Only if the user selects **"Proceed"** or **"Start"** will Alfred call the tdd-i
 
 ## 🧠 Context Management
 
-> For more information: `.moai/memory/development-guide.md` - see section "Context Engineering"
+> For more information: Skill("moai-alfred-dev-guide") - see section "Context Engineering"
 
 ### Core strategy of this command
 
@@ -536,6 +661,45 @@ Only if the user selects **"Proceed"** or **"Start"** will Alfred call the tdd-i
 
 ---
 
+## Final Step
+
+### After STEP 3 (git-manager) Completes
+
+Alfred calls AskUserQuestion to collect user's next action:
+
+```python
+AskUserQuestion(
+    questions=[
+        {
+            "question": "구현이 완료되었습니다. 다음으로 뭘 하시겠습니까?",
+            "header": "다음 단계",
+            "multiSelect": false,
+            "options": [
+                {
+                    "label": "📚 문서 동기화 진행",
+                    "description": "/alfred:3-sync 실행하여 문서 동기화"
+                },
+                {
+                    "label": "🔍 추가 구현",
+                    "description": "다른 SPEC 구현 진행"
+                },
+                {
+                    "label": "🔄 새 세션 시작",
+                    "description": "성능 최적화를 위해 /clear 실행"
+                }
+            ]
+        }
+    ]
+)
+```
+
+**User Responses**:
+- **📚 문서 동기화**: Proceed to `/alfred:3-sync` for documentation synchronization
+- **🔍 추가 구현**: Repeat `/alfred:2-run SPEC-XXX` for next feature
+- **🔄 새 세션**: Execute `/clear` to start fresh session (recommended for performance)
+
+---
+
 ## Next steps
 
 **Recommendation**: For better performance and context management, start a new chat session with the `/clear` or `/new` command before proceeding to the next step.