musubi-sdd 5.9.6 → 6.1.0

package/README.ja.md

@@ -262,10 +262,10 @@ musubi init -r owner/repo@develop
 
 MUSUBIは7つのAIコーディングエージェントに対応し、それぞれに最適化された設定を提供します。
 
-| エージェント       | スキルAPI     | 27スキル | コマンド形式     | コマンドファイル形式 | インストールディレクトリ                      |
-| ------------------ | ------------- | -------------- | ---------------- | -------------------- | --------------------------------------------- |
-| **Claude Code**    | ✅ (27スキル) | ✅             | `/sdd-*`         | Markdown             | `.claude/skills/`, `.claude/commands/`        |
-| **GitHub Copilot** | ❌            | ✅ (AGENTS.md) | `#sdd-*`         | Markdown + AGENTS.md | `.github/prompts/`, `.github/AGENTS.md`       |
+| エージェント       | スキルAPI     | 27スキル | コマンド形式     | コマンドファイル形式       | インストールディレクトリ                      |
+| ------------------ | ------------- | -------------- | ---------------- | ------------------------- | --------------------------------------------- |
+| **Claude Code**    | ✅ (27スキル) | ✅             | `/sdd-*`         | Markdown (.md)            | `.claude/skills/`, `.claude/commands/`        |
+| **GitHub Copilot** | ❌            | ✅ (AGENTS.md) | `/sdd-*`         | プロンプトファイル (.prompt.md) | `.github/prompts/`, `.github/AGENTS.md`       |
 | **Cursor IDE**     | ❌            | ✅ (AGENTS.md) | `/sdd-*`         | Markdown + AGENTS.md | `.cursor/commands/`, `.cursor/AGENTS.md`      |
 | **Gemini CLI**     | ❌            | ✅ (GEMINI.md) | `/sdd-*`         | TOML + GEMINI.md     | `.gemini/commands/`, `GEMINI.md`              |
 | **Codex CLI**      | ❌            | ✅ (AGENTS.md) | `/prompts:sdd-*` | Markdown + AGENTS.md | `.codex/prompts/`, `.codex/AGENTS.md`         |
@@ -277,6 +277,7 @@ MUSUBIは7つのAIコーディングエージェントに対応し、それぞ
 - スキルAPIはClaude Code専用です
 - **全7プラットフォームが27スキルに対応**（Skills APIまたはAGENTS.md経由）
 - v5.9.0で5つの組み込みオーケストレータースキルを追加（release, workflow, package, constitution-level, project-config）
+- **GitHub CopilotはVS Code公式ドキュメントに従い `.prompt.md` 拡張子を使用**
 - AGENTS.md: OpenAI仕様、GitHub Copilot & Cursorが公式サポート
 - Gemini CLIはTOML形式 + GEMINI.md統合を使用
 - その他のエージェントはMarkdown形式 + AGENTS.mdを使用
@@ -535,8 +536,11 @@ your-project/
 
 ```text
 your-project/
-├── .github/prompts/         # GitHub Copilot用（#sdd-*、Markdown）
-│   ├── AGENTS.md             # 25エージェント定義（公式サポート）
+├── .github/prompts/         # GitHub Copilot用（/sdd-*、.prompt.md）
+│   ├── sdd-steering.prompt.md    # プロンプトファイルは .prompt.md 拡張子
+│   ├── sdd-requirements.prompt.md
+│   ├── ... (9つのプロンプトファイル)
+│   └── AGENTS.md             # 27エージェント定義（公式サポート）
 │   または
 ├── .cursor/commands/        # Cursor用（/sdd-*、Markdown）
 │   ├── AGENTS.md             # 25エージェント定義（公式サポート）
@@ -563,11 +567,12 @@ your-project/
 
 **主な違い**：
 
-- **Claude Code**: 25 Skills API（専用） + コマンド（Markdown）
-- **GitHub Copilot & Cursor**: AGENTS.md（公式サポート） + コマンド（Markdown）
-- **Gemini CLI**: GEMINI.md統合（25エージェント） + TOMLコマンド（ユニーク）
+- **Claude Code**: 27 Skills API（専用） + コマンド（Markdown）
+- **GitHub Copilot**: AGENTS.md（公式サポート） + プロンプトファイル (`.prompt.md`)
+- **Cursor**: AGENTS.md（公式サポート） + コマンド（Markdown）
+- **Gemini CLI**: GEMINI.md統合（27エージェント） + TOMLコマンド（ユニーク）
 - **その他**: AGENTS.md（互換） + Markdownコマンド
-- **全プラットフォーム**: 同じ25エージェント、異なる実装形式
+- **全プラットフォーム**: 同じ27エージェント、異なる実装形式
 
 ## 使用方法
 

package/README.md

@@ -365,10 +365,10 @@ const result = await validator.validateAll(projectPath);
 
 MUSUBI supports 7 AI coding agents, each with tailored configurations:
 
-| Agent              | Skills API     | 27 Skills      | Command Format   | Command File Format  | Installation Directory                        |
-| ------------------ | -------------- | -------------- | ---------------- | -------------------- | --------------------------------------------- |
-| **Claude Code**    | ✅ (27 skills) | ✅             | `/sdd-*`         | Markdown             | `.claude/skills/`, `.claude/commands/`        |
-| **GitHub Copilot** | ❌             | ✅ (AGENTS.md) | `#sdd-*`         | Markdown + AGENTS.md | `.github/prompts/`, `.github/AGENTS.md`       |
+| Agent              | Skills API     | 27 Skills      | Command Format   | Command File Format       | Installation Directory                        |
+| ------------------ | -------------- | -------------- | ---------------- | ------------------------- | --------------------------------------------- |
+| **Claude Code**    | ✅ (27 skills) | ✅             | `/sdd-*`         | Markdown (.md)            | `.claude/skills/`, `.claude/commands/`        |
+| **GitHub Copilot** | ❌             | ✅ (AGENTS.md) | `/sdd-*`         | Prompt File (.prompt.md)  | `.github/prompts/`, `.github/AGENTS.md`       |
 | **Cursor IDE**     | ❌             | ✅ (AGENTS.md) | `/sdd-*`         | Markdown + AGENTS.md | `.cursor/commands/`, `.cursor/AGENTS.md`      |
 | **Gemini CLI**     | ❌             | ✅ (GEMINI.md) | `/sdd-*`         | TOML + GEMINI.md     | `.gemini/commands/`, `GEMINI.md`              |
 | **Codex CLI**      | ❌             | ✅ (AGENTS.md) | `/prompts:sdd-*` | Markdown + AGENTS.md | `.codex/prompts/`, `.codex/AGENTS.md`         |
@@ -380,6 +380,7 @@ MUSUBI supports 7 AI coding agents, each with tailored configurations:
 - Skills API is exclusive to Claude Code
 - **All 7 platforms now support 27 skills** via Skills API (Claude Code) or AGENTS.md (others)
 - v5.9.0 added 5 built-in orchestrator skills (release, workflow, package, constitution-level, project-config)
+- **GitHub Copilot uses `.prompt.md` extension** per VS Code official documentation
 - AGENTS.md: OpenAI specification, officially supported by GitHub Copilot & Cursor
 - Gemini CLI uses TOML format + GEMINI.md integration
 
@@ -646,8 +647,11 @@ your-project/
 
 ```text
 your-project/
-├── .github/prompts/         # For GitHub Copilot (#sdd-*, Markdown)
-│   ├── AGENTS.md             # 27 skills definition (official support)
+├── .github/prompts/         # For GitHub Copilot (/sdd-*, .prompt.md)
+│   ├── sdd-steering.prompt.md    # Prompt files use .prompt.md extension
+│   ├── sdd-requirements.prompt.md
+│   ├── ... (9 prompt files)
+│   └── AGENTS.md             # 27 skills definition (official support)
 │   OR
 ├── .cursor/commands/        # For Cursor (/sdd-*, Markdown)
 │   ├── AGENTS.md             # 27 skills definition (official support)
@@ -684,7 +688,8 @@ your-project/
 **Key Differences**:
 
 - **Claude Code**: 27 Skills (25 + 5 built-in) + commands (Markdown)
-- **GitHub Copilot & Cursor**: AGENTS.md (official support) + commands (Markdown)
+- **GitHub Copilot**: AGENTS.md (official support) + prompt files (`.prompt.md`)
+- **Cursor**: AGENTS.md (official support) + commands (Markdown)
 - **Gemini CLI**: GEMINI.md integration (27 skills) + TOML commands (unique)
 - **Others**: AGENTS.md (compatible) + Markdown commands
 - **All platforms**: Same 27 skills, different implementation formats

package/bin/musubi-init.js

@@ -1253,7 +1253,7 @@ async function main(agent, agentKey, options = {}) {
     'templates',
     'storage/specs',
     'storage/changes',
-    'storage/features',
+    'storage/archive',
   ];
 
   // Add agent-specific directories

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "musubi-sdd",
-  "version": "5.9.6",
+  "version": "6.1.0",
   "description": "Ultimate Specification Driven Development Tool with 27 Agents for 7 AI Coding Platforms + MCP Integration (Claude Code, GitHub Copilot, Cursor, Gemini CLI, Windsurf, Codex, Qwen Code)",
   "main": "src/index.js",
   "bin": {

package/src/templates/agents/claude-code/skills/constitution-enforcer/SKILL.md

@@ -616,4 +616,4 @@ Before finishing:
 - [ ] Failures documented with severity
 - [ ] Remediation plan provided
 - [ ] Overall status determined (APPROVED/BLOCKED)
-- [ ] Report saved to `storage/features/[feature]/constitutional-compliance.md`
+- [ ] Report saved to `storage/specs/[feature]/constitutional-compliance.md`

package/src/templates/agents/claude-code/skills/constitution-enforcer/phase-minus-one-gates.md

@@ -192,9 +192,9 @@ def validate_traceability_setup(feature_name):
     Verify traceability chain is ready.
     """
     required_artifacts = {
-        "requirements": f"storage/features/{feature_name}/requirements.md",
-        "design": f"storage/features/{feature_name}/design.md",
-        "tasks": f"storage/features/{feature_name}/tasks.md",
+        "requirements": f"storage/specs/{feature_name}/requirements.md",
+        "design": f"storage/specs/{feature_name}/design.md",
+        "tasks": f"storage/specs/{feature_name}/tasks.md",
     }
 
     for artifact, path in required_artifacts.items():

package/src/templates/agents/claude-code/skills/site-reliability-engineer/SKILL.md

@@ -318,7 +318,7 @@ app.get('/health/live', (req, res) => {
 
 ### Phase 1: SLO Definition (Based on Requirements)
 
-1. Read `storage/features/[feature]/requirements.md`
+1. Read `storage/specs/[feature]-requirements.md`
 2. Identify non-functional requirements (performance, availability)
 3. Define SLIs and SLOs
 4. Calculate error budgets

package/src/templates/agents/claude-code/skills/steering/auto-update-rules.md

@@ -107,7 +107,7 @@ git diff --name-only | grep -E "(webpack|vite|rollup|esbuild)\.config\."
 
 ```bash
 # New feature directories
-git diff --name-only --diff-filter=A | grep "storage/features/"
+git diff --name-only --diff-filter=A | grep "storage/specs/"
 
 # New requirement files
 git diff --name-only --diff-filter=A | grep "requirements\.md"
@@ -154,7 +154,7 @@ git diff --name-only --diff-filter=A | grep "requirements\.md"
 
 ### memories/architecture_decisions.md
 
-**Trigger**: New ADR created in `storage/features/*/design.md`
+**Trigger**: New ADR created in `storage/specs/*/design.md`
 
 **Action**:
 
@@ -245,8 +245,8 @@ diff <(grep "dependencies" steering/tech.md) <(cat package.json | jq '.dependenc
 # 2. Check if structure.md directories match actual structure
 diff <(grep -E "^-\s+" steering/structure.md | cut -d' ' -f2) <(find src -type d -maxdepth 2)
 
-# 3. Check if product.md features match storage/features/
-diff <(grep "Feature" steering/product.md) <(ls storage/features/)
+# 3. Check if product.md features match storage/specs/
+diff <(grep "Feature" steering/product.md) <(ls storage/specs/)
 ```
 
 ### Report Divergence
@@ -268,7 +268,7 @@ diff <(grep "Feature" steering/product.md) <(ls storage/features/)
 
 ### product.md
 
-- ⚠️ Missing feature: payment-integration (in storage/features/)
+- ⚠️ Missing feature: payment-integration (in storage/specs/)
 
 ## Recommendations
 
@@ -310,7 +310,7 @@ on:
     paths:
       - 'package.json'
       - 'src/**'
-      - 'storage/features/**'
+      - 'storage/specs/**'
 
 jobs:
   check-steering:

package/src/templates/agents/claude-code/skills/traceability-auditor/SKILL.md

@@ -101,9 +101,9 @@ Test Suite (tests/auth/service.test.ts)
 
 ### Phase 1: Collect Artifacts
 
-1. Read `storage/features/[feature]/requirements.md`
-2. Read `storage/features/[feature]/design.md`
-3. Read `storage/features/[feature]/tasks.md`
+1. Read `storage/specs/[feature]-requirements.md`
+2. Read `storage/specs/[feature]-design.md`
+3. Read `storage/specs/[feature]-tasks.md`
 4. Scan source code for implementation
 5. Scan test files for test cases
 
@@ -290,7 +290,7 @@ coverage_test = (requirements_with_tests / requirements_total) * 100
 - **After**:
   - If gaps found → orchestrator triggers missing skills
   - If complete → quality-assurance approves release
-- **Uses**: All spec files in `storage/features/` and `storage/changes/`
+- **Uses**: All spec files in `storage/specs/` and `storage/changes/`
 
 ## Gap Detection Rules
 

package/src/templates/agents/claude-code/skills/traceability-auditor/coverage-matrix-template.md

@@ -99,9 +99,9 @@ Template for creating traceability coverage matrices that ensure 100% compliance
 def generate_traceability_matrix(feature_name):
     """Generate traceability matrix for a feature."""
 
-    requirements = parse_requirements(f"storage/features/{feature_name}/requirements.md")
-    design = parse_design(f"storage/features/{feature_name}/design.md")
-    tasks = parse_tasks(f"storage/features/{feature_name}/tasks.md")
+    requirements = parse_requirements(f"storage/specs/{feature_name}-requirements.md")
+    design = parse_design(f"storage/specs/{feature_name}-design.md")
+    tasks = parse_tasks(f"storage/specs/{feature_name}-tasks.md")
     code_files = find_code_files(f"src/{feature_name}/")
     test_files = find_test_files(f"tests/{feature_name}/")
 

package/src/templates/agents/claude-code/skills/traceability-auditor/gap-detection-rules.md

@@ -94,8 +94,8 @@ for ref in all_references:
 #!/bin/bash
 # find-orphaned-requirements.sh
 
-REQ_FILE="storage/features/$1/requirements.md"
-DESIGN_FILE="storage/features/$1/design.md"
+REQ_FILE="storage/specs/$1-requirements.md"
+DESIGN_FILE="storage/specs/$1-design.md"
 
 # Extract requirement IDs
 grep -oP 'REQ-[\w-]+' "$REQ_FILE" | sort -u > /tmp/reqs.txt
@@ -212,7 +212,7 @@ name: Traceability Check
 on:
   pull_request:
     paths:
-      - 'storage/features/**'
+      - 'storage/specs/**'
       - 'src/**'
       - 'tests/**'
 

package/src/templates/agents/github-copilot/commands/{sdd-requirements.md → sdd-requirements.prompt.md}

@@ -1,10 +1,10 @@
 # SDD Requirements Command
 
-Create EARS-format requirements specification.
+Create EARS-format requirements specification through interactive dialogue.
 
 ---
 
-## Instructions for Claude
+## Instructions for AI Agent
 
 You are executing the `/sdd-requirements [feature-name]` command to create a requirements specification.
 
@@ -18,7 +18,14 @@ You are executing the `/sdd-requirements [feature-name]` command to create a req
 
 ### Your Task
 
-Generate a comprehensive requirements specification in EARS format for the specified feature.
+**CRITICAL**: Before generating requirements, you MUST engage in an interactive 1-on-1 dialogue with the user to uncover the TRUE PURPOSE behind their request. Surface-level requirements often hide deeper needs.
+
+### Output Directory
+
+**Requirements documents are saved to**: `storage/specs/`
+
+- English: `storage/specs/{{feature-name}}-requirements.md`
+- Japanese: `storage/specs/{{feature-name}}-requirements.ja.md`
 
 ---
 
@@ -44,28 +51,103 @@ steering/tech.md         # Technology stack
 
 ---
 
-### 2. Gather Requirements
+### 2. Interactive True Purpose Discovery (1-on-1 Dialogue)
+
+**CRITICAL**: Conduct a 1-question-at-a-time dialogue to uncover the TRUE PURPOSE.
+
+#### Phase 1: Surface Purpose → True Purpose Discovery
+
+**Ask ONE question at a time. Wait for the user's response before proceeding.**
+
+**Question Flow**:
 
-**Methods** (use as appropriate):
+1. **Why Question** (Dig deeper into purpose)
+   - "この機能で解決したい『本当の課題』は何ですか？"
+   - "この機能がないと、誰がどのように困りますか？"
 
-#### A. Stakeholder Interview (if user is available)
+2. **Who Question** (Identify stakeholders)
+   - "この機能を最も必要としているのは誰ですか？"
+   - "この機能の成功を判断する人は誰ですか？"
 
-Use `AskUserQuestion` tool to ask:
+3. **What-If Question** (Validate assumptions)
+   - "もしこの機能が完璧に動作したら、何が変わりますか？"
+   - "この機能なしで、代替手段はありますか？"
 
-- Who are the users of this feature?
-- What problem does this solve?
-- What are the critical workflows?
-- What are the acceptance criteria?
-- Are there any constraints (performance, security, compliance)?
+4. **Constraint Question** (Uncover hidden limits)
+   - "この機能で『絶対にやってはいけないこと』はありますか？"
+   - "技術的・ビジネス的な制約はありますか？"
+
+5. **Success Criteria Question** (Define done)
+   - "この機能が『成功した』と言えるのはどんな状態ですか？"
+   - "具体的な数値目標（KPI）はありますか？"
+
+**IMPORTANT**:
+- Ask only ONE question and wait for response
+- Based on the answer, ask follow-up questions
+- Continue until the TRUE PURPOSE is clear
+- Summarize understanding and confirm with user
+
+---
+
+### 3. MECE Analysis (Comprehensive Coverage)
+
+**After understanding the TRUE PURPOSE, apply MECE to ensure completeness.**
+
+#### MECE Dimensions for Requirements
+
+**Dimension 1: User Journey (Who × When × Where)**
+| Stage | Questions to Cover |
+|-------|--------------------|
+| Before | How do users discover/access this feature? |
+| During | What is the core interaction flow? |
+| After | What happens after completion? Notifications? |
+| Error | What happens when something goes wrong? |
+
+**Dimension 2: Data Lifecycle (CRUD)**
+| Operation | Check |
+|-----------|-------|
+| Create | What data is created? By whom? |
+| Read | Who can view? What format? |
+| Update | What can be modified? By whom? |
+| Delete | Soft delete? Hard delete? Retention? |
+
+**Dimension 3: Cross-Cutting Concerns**
+| Concern | Requirements |
+|---------|-------------|
+| Security | Authentication, Authorization, Encryption |
+| Performance | Response time, Throughput, Concurrency |
+| Reliability | Error handling, Recovery, Failover |
+| Scalability | Load limits, Growth capacity |
+| Compliance | GDPR, SOC2, PCI-DSS (if applicable) |
+
+**Dimension 4: Integration Points**
+| Integration | Check |
+|-------------|-------|
+| Upstream | What systems feed data to this feature? |
+| Downstream | What systems consume this feature's output? |
+| External | Any third-party APIs or services? |
+
+**MECE Completeness Check**:
+- [ ] All user types covered?
+- [ ] All states (normal, edge, error) covered?
+- [ ] All data flows covered?
+- [ ] All integration points covered?
+- [ ] No overlapping requirements?
+
+---
 
-#### B. Research Existing System (brownfield)
+### 4. Research Existing System (Brownfield)
+
+**If modifying existing functionality**:
 
 - Search for existing implementation: `grep -r "{{feature}}" src/`
 - Read related code
 - Identify current behavior
 - Document what needs to change (delta spec)
 
-#### C. Infer from Context
+---
+
+### 5. Contextual Research
 
 - Analyze steering/product.md for user types
 - Review existing requirements docs
@@ -73,7 +155,9 @@ Use `AskUserQuestion` tool to ask:
 
 ---
 
-### 3. Generate Requirements Document
+### 6. Generate Requirements Document
+
+**Output Directory**: `storage/specs/`
 
 Use template from `templates/requirements.md`:
 
@@ -84,13 +168,41 @@ Use template from `templates/requirements.md`:
 
 ## Overview
 
-- Purpose
+- Purpose (True Purpose discovered through dialogue)
 - Scope (in/out)
 - Business context (from steering/product.md)
 
+## True Purpose Statement
+
+[Summary of TRUE PURPOSE uncovered through dialogue]
+
+- Surface Request: [What user initially asked for]
+- True Purpose: [What user actually needs]
+- Key Insight: [Critical understanding gained]
+
 ## Stakeholders
 
-[Table of roles]
+[Table of roles identified through dialogue]
+
+## MECE Coverage Summary
+
+### User Journey Coverage
+- [ ] Before: Entry/Discovery
+- [ ] During: Core Flow
+- [ ] After: Completion
+- [ ] Error: Exception Handling
+
+### Data Lifecycle Coverage
+- [ ] Create
+- [ ] Read
+- [ ] Update
+- [ ] Delete
+
+### Cross-Cutting Concerns
+- [ ] Security
+- [ ] Performance
+- [ ] Reliability
+- [ ] Scalability
 
 ## Functional Requirements
 
@@ -98,6 +210,8 @@ Use template from `templates/requirements.md`:
 
 [EARS Pattern - choose appropriate pattern]
 
+**MECE Category**: [User Journey Stage | Data Operation | etc.]
+
 **Acceptance Criteria**:
 
 - [Testable criterion 1]
@@ -119,6 +233,10 @@ Use template from `templates/requirements.md`:
 
 ### REQ-AVAIL-001: Availability
 
+## MECE Gap Analysis
+
+[List any gaps identified during MECE analysis]
+
 ## Requirements Coverage Matrix
 
 [Initial table - will be filled during design/implementation]
@@ -126,7 +244,7 @@ Use template from `templates/requirements.md`:
 
 ---
 
-### 4. Apply EARS Format (Article IV)
+### 7. Apply EARS Format (Article IV)
 
 **CRITICAL**: All requirements MUST use one of 5 EARS patterns.
 
@@ -150,10 +268,11 @@ Each requirement MUST have:
 - [ ] Testable acceptance criteria
 - [ ] Priority (P0/P1/P2/P3)
 - [ ] Status (Draft initially)
+- [ ] MECE Category (for traceability)
 
 ---
 
-### 5. Assign Requirement IDs
+### 8. Assign Requirement IDs
 
 **Format**: `REQ-[COMPONENT]-[NUMBER]`
 
@@ -172,7 +291,7 @@ Each requirement MUST have:
 
 ---
 
-### 6. Add Non-Functional Requirements
+### 9. Add Non-Functional Requirements (from MECE Cross-Cutting Concerns)
 
 Always include these categories:
 
@@ -221,7 +340,7 @@ The {{COMPONENT}} SHALL maintain 99.9% uptime.
 
 ---
 
-### 7. Brownfield: Create Delta Specification
+### 10. Brownfield: Create Delta Specification
 
 If this is a change to existing system, add delta sections:
 
@@ -266,7 +385,7 @@ The authentication system SHALL hash passwords using bcrypt cost 12.
 
 ---
 
-### 8. Constitutional Validation
+### 11. Constitutional Validation
 
 Validate requirements against constitutional articles:
 
@@ -290,10 +409,12 @@ Run validation:
 
 ---
 
-### 9. Save Document (Bilingual)
+### 12. Save Document (Bilingual)
 
 **IMPORTANT**: Create BOTH English and Japanese versions.
 
+**Output Directory**: `storage/specs/`
+
 **English version (Primary/Reference)**:
 Save to: `storage/specs/{{feature-name}}-requirements.md`
 
@@ -323,7 +444,7 @@ Save to: `storage/specs/{{feature-name}}-requirements.ja.md`
 
 ---
 
-### 10. Generate Summary
+### 13. Generate Summary
 
 Present summary to user:
 
@@ -331,11 +452,25 @@ Present summary to user:
 ## ✅ Requirements Specification Complete
 
 **Feature**: {{FEATURE_NAME}}
+**Output Directory**: `storage/specs/`
 **Files**:
 
 - English: storage/specs/{{feature-name}}-requirements.md
 - Japanese: storage/specs/{{feature-name}}-requirements.ja.md
 
+### True Purpose Discovery:
+
+- Surface Request: [What user initially asked]
+- True Purpose: [What user actually needs]
+- Key Insight: [Critical understanding gained]
+
+### MECE Coverage:
+
+- User Journey: ✅ All stages covered
+- Data Lifecycle: ✅ CRUD covered
+- Cross-Cutting: ✅ Security/Performance/Reliability/Scalability
+- Integrations: ✅ Upstream/Downstream/External
+
 ### Summary:
 
 - **Total Requirements**: [N]
@@ -381,28 +516,54 @@ Present summary to user:
 ### User Command:
 
 ```bash
-/sdd-requirements authentication
+#sdd-requirements authentication
 ```
 
-### Claude Actions:
+### AI Agent Actions:
 
 1. **Read Steering**:
    - `steering/product.md` → Target users: B2B SaaS companies
    - `steering/tech.md` → Stack: Next.js, PostgreSQL, Prisma
    - `steering/structure.md` → Pattern: Library-first
 
-2. **Ask User** (if needed):
+2. **Interactive Dialogue (1-on-1)**:
 
-   ```
-   Questions about authentication feature:
-   1. Authentication methods: Email/password, OAuth, SSO, or all?
-   2. Required features: 2FA, password reset, session management?
-   3. Security: Any compliance requirements (SOC2, HIPAA)?
-   ```
+   **AI**: 「認証機能について、この機能で解決したい『本当の課題』は何ですか？」
+
+   **User**: 「ユーザーがログインできるようにしたい」
+
+   **AI**: 「なるほど。では、この機能がないと、誰がどのように困りますか？」
+
+   **User**: 「顧客がサービスを使えない。また、管理者が顧客データを保護できない」
+
+   **AI**: 「顧客データの保護が重要なのですね。この機能で『絶対にやってはいけないこと』はありますか？」
 
-3. **Generate Requirements**:
+   **User**: 「データ漏洩。SOC2準拠が必要」
+
+   **AI**: 「SOC2準拠が必要とのこと。この機能が『成功した』と言えるのはどんな状態ですか？」
+
+   **User**: 「ログイン成功率99%以上、不正アクセスゼロ」
+
+   **AI**: 「理解しました。確認させてください：
+   - **表面的要求**: ユーザーがログインできること
+   - **真の目的**: SOC2準拠でセキュアな顧客データ保護
+   - **成功基準**: ログイン成功率99%以上、不正アクセスゼロ
+   この理解で正しいですか？」
+
+3. **MECE Analysis**:
+   - User Journey: Login → Session → Logout → Error handling
+   - Data: Credentials(C), Session(RUD), Audit Log(CR)
+   - Security: Authentication, Authorization, Encryption, Audit
+
+4. **Generate Requirements**:
 
    ```markdown
+   ## True Purpose Statement
+
+   - Surface Request: ユーザーログイン機能
+   - True Purpose: SOC2準拠のセキュアな顧客データ保護基盤
+   - Key Insight: 単なるログインではなく、コンプライアンス要件が重要
+
    ### REQ-AUTH-001: User Login
 
    WHEN a user provides valid credentials,
@@ -410,19 +571,19 @@ Present summary to user:
    AND the system SHALL create a session
    AND the system SHALL redirect to dashboard.
 
-   ### REQ-AUTH-002: Password Strength
+   **MECE Category**: User Journey - During
 
-   The authentication system SHALL enforce password requirements:
+   ### REQ-AUTH-002: Audit Logging (SOC2)
 
-   - Minimum 12 characters
-   - At least 1 uppercase, 1 lowercase, 1 number, 1 special char
+   The authentication system SHALL log all authentication events
+   with timestamp, user ID, IP address, and result.
 
-   [... more requirements ...]
+   **MECE Category**: Cross-Cutting - Security/Compliance
    ```
 
-4. **Save**: `storage/specs/authentication-requirements.md`
+5. **Save to Output Directory**: `storage/specs/authentication-requirements.md`
 
-5. **Summarize**: Present summary to user
+6. **Summarize**: Present summary with MECE coverage report
 
 ---
 
@@ -446,15 +607,18 @@ Present summary to user:
 
 Before completing, verify:
 
+- [ ] Interactive dialogue conducted (TRUE PURPOSE discovered)
+- [ ] MECE analysis completed (all dimensions covered)
 - [ ] Steering context read and applied
 - [ ] All requirements in EARS format
 - [ ] Requirement IDs assigned (REQ-XXX-NNN)
+- [ ] MECE category assigned to each requirement
 - [ ] Acceptance criteria defined for each requirement
 - [ ] Priority assigned (P0/P1/P2/P3)
 - [ ] Non-functional requirements included
 - [ ] Constitutional validation passed
 - [ ] Document saved to storage/specs/
-- [ ] Summary presented to user
+- [ ] Summary with MECE coverage presented to user
 
 ---
 
@@ -486,4 +650,14 @@ If `storage/specs/{{feature-name}}-requirements.md` exists:
 
 ---
 
+**Output Directory Summary**:
+- Requirements documents: `storage/specs/{{feature-name}}-requirements.md`
+- Japanese version: `storage/specs/{{feature-name}}-requirements.ja.md`
+
+---
+
+**Execution**: Begin interactive dialogue now for the specified feature.
+
+---
+
 **Execution**: Begin requirements generation now for the specified feature.