{
"markdown": "\n\n# π MUSUBI\n\n**Ultimate Specification Driven Development Tool**\n\n[](https://github.com/nahisaho/MUSUBI/actions/workflows/ci.yml)\n[](https://www.npmjs.com/package/musubi-sdd)\n[](https://www.npmjs.com/package/musubi-sdd)\n[](https://opensource.org/licenses/MIT)\n[](https://nodejs.org)\n[](http://makeapullrequest.com)\n\n[English](README.md) | [ζ₯ζ¬θͺ](README.ja.md) | [Documentation](docs/) | [npm](https://www.npmjs.com/package/musubi-sdd)\n\n
\n\n---\n\n> π€ **7 AI Coding Agents** Γ π **31 Specialized Skills** Γ βοΈ **Constitutional Governance**\n\nMUSUBI (η΅γ³ - \"connection/binding\") is a comprehensive **Specification Driven Development (SDD)** framework that synthesizes the best features from 6 leading frameworks into a production-ready tool for multiple AI coding agents.\n\n\n\n\n\n
\n\n## β¨ Why MUSUBI?\n\n| Challenge | MUSUBI Solution |\n| ------------------------- | ---------------------------------------------- |\n| π Fragmented AI tools | **7 agents, 1 unified workflow** |\n| π Ambiguous requirements | **EARS format with 5 patterns** |\n| π Lost traceability | **100% ReqβDesignβCodeβTest mapping** |\n| β οΈ Quality inconsistency | **9 Constitutional Articles + Phase -1 Gates** |\n| π Brownfield challenges | **Delta specs + Change management** |\n\n## π Quick Start\n\n```bash\n# Install and initialize in 30 seconds\nnpx musubi-sdd init\n\n# For existing projects - auto-analyze and generate steering docs\nnpx musubi-sdd onboard\n\n# That's it! Now use your AI agent with SDD commands:\n# Claude Code: /sdd-requirements, /sdd-design, /sdd-implement\n# GitHub Copilot: #sdd-requirements, #sdd-design, #sdd-implement\n```\n\n\nπ¦ More installation options
\n\n```bash\n# Global installation\nnpm install -g musubi-sdd\n\n# Initialize for specific AI agent\nmusubi init --copilot # GitHub Copilot\nmusubi init --cursor # Cursor IDE\nmusubi init --gemini # Gemini CLI\nmusubi init --codex # Codex CLI\nmusubi init --qwen # Qwen Code\nmusubi init --windsurf # Windsurf IDE\n```\n\n \n\n---\n\n## π What's New in v6.3.0\n\n### SDD Document Path Unification π\n\nComplete separation of document types for better organization and clarity.\n\n#### Storage Structure\n\n| Document Type | Storage Path | Naming Pattern |\n|---------------|--------------|----------------|\n| **Requirements** | `storage/specs/` | `{feature}-requirements.md` |\n| **Design** | `storage/design/` | `{feature}-design.md` |\n| **Tasks** | `storage/tasks/` | `{feature}-tasks.md` |\n| **Validation** | `storage/validation/` | `{feature}-validation-report.md` |\n\n```\nstorage/\nβββ specs/ # Requirements only\nβ βββ auth-requirements.md\nβββ design/ # Design documents only\nβ βββ auth-design.md\nβββ tasks/ # Tasks only\nβ βββ auth-tasks.md\nβββ validation/ # Validation reports\n βββ auth-validation-report.md\n```\n\n#### Updated All 7 Agent Platforms\n\nAll prompts and templates updated with correct storage paths:\n- Claude Code, GitHub Copilot, Cursor, Gemini CLI, Codex CLI, Qwen Code, Windsurf\n\n#### Test Results\n\n- **Unit Tests**: 4,827 passing β
\n- **Integration Tests**: 660 passing β
\n\n---\n\n## π What's New in v6.2.0\n\n### Review Gate Engine π‘οΈ\n\nNew quality gates for systematic review at each development stage.\n\n#### Review Gates\n\n| Gate | Description | Prompt |\n|------|-------------|--------|\n| **Requirements Gate** | EARS format, priority, acceptance criteria validation | `#sdd-review-requirements` |\n| **Design Gate** | C4 model, ADR, component design validation | `#sdd-review-design` |\n| **Implementation Gate** | Code quality, test coverage, naming conventions | `#sdd-review-implementation` |\n| **Full Review** | All gates in sequence | `#sdd-review-all` |\n\n```bash\n# Use review prompts in your AI agent\n#sdd-review-requirements user-auth\n#sdd-review-design user-auth\n#sdd-review-implementation user-auth\n```\n\n#### Workflow Dashboard\n\n- **Progress Visualization**: Real-time progress across 5 stages\n- **Blocker Management**: Add, resolve, and track blockers\n- **Transition Recording**: Record and analyze stage transitions\n- **Sprint Planning**: Task prioritization with velocity tracking\n\n```bash\nmusubi dashboard show # Show workflow status\nmusubi dashboard start # Start new workflow\nmusubi dashboard blocker # Manage blockers\n```\n\n#### Traceability System\n\n- **Auto-Extraction**: Automatic ID extraction from code, tests, commits\n- **Gap Detection**: Detect missing design, implementation, or tests\n- **Matrix Storage**: YAML-based traceability matrix with history\n\n```bash\nmusubi-trace extract # Extract traceability IDs\nmusubi-trace gaps # Detect gaps\nmusubi-trace matrix # Generate matrix report\n```\n\n#### Enterprise Features\n\n| Feature | Description |\n|---------|-------------|\n| **Error Recovery** | Auto error analysis with remediation steps |\n| **Rollback Manager** | File/commit/stage/sprint level rollback |\n| **CI Reporter** | GitHub Actions integration |\n| **Phase -1 Gate** | Constitutional compliance before all stages |\n| **Steering Sync** | Auto-update steering files on version change |\n\n---\n\n## π What's New in v5.9.0\n\n### Phase 1-4 Enterprise Features π’\n\nMajor update with enterprise-ready features for large-scale projects and monorepo support.\n\n#### Workflow Flexibility (Phase 1)\n\n- **3 Workflow Modes**: `small` (bug fixes), `medium` (features), `large` (architecture)\n- **Auto-detection**: Smart mode selection based on feature name patterns\n- **`musubi-release`**: New CLI for release automation with CHANGELOG generation\n\n```bash\n# Generate CHANGELOG from commits\nmusubi-release\n\n# Show detected mode for a feature\nmusubi-workflow mode --detect \"feat: user authentication\"\n```\n\n#### Monorepo Support (Phase 2)\n\n- **Package Registry**: `steering/packages.yml` for dependency management\n- **Dependency Graphs**: Mermaid diagram generation for visualization\n- **Coverage Tracking**: Per-package test coverage reporting\n\n#### Constitution Level Management (Phase 3)\n\n- **3 Enforcement Levels**: `critical` (blocking), `advisory` (warnings), `flexible` (suggestions)\n- **Level-aware Validation**: Different enforcement based on article importance\n- **Project Overrides**: Custom levels per project type\n\n| Level | Articles | Behavior |\n|-------|----------|----------|\n| Critical | CONST-001, 002, 003, 005, 009 | Blocks workflow |\n| Advisory | CONST-004, 006, 007 | Warnings only |\n| Flexible | CONST-008 | Suggestions |\n\n#### Project Configuration (Phase 4)\n\n- **`musubi-config`**: New CLI for configuration management\n- **Schema Validation**: v2.0 schema with AJV validation\n- **Auto-migration**: Upgrade v1.0 configs to v2.0\n\n```bash\nmusubi-config validate # Validate project.yml\nmusubi-config migrate # Migrate to v2.0\nmusubi-config show # Show effective config\n```\n\n#### Orchestrator Integration\n\n5 new built-in skills for programmatic access:\n\n| Skill | Category | Usage |\n|-------|----------|-------|\n| `release-manager` | release | CHANGELOG generation |\n| `workflow-mode-manager` | workflow | Mode detection & management |\n| `package-manager` | configuration | Package & dependency analysis |\n| `constitution-level-manager` | validation | Level-aware validation |\n| `project-config-manager` | configuration | Config validation & migration |\n\n```javascript\nconst { workflowModeSkill } = require('musubi-sdd/src/orchestration');\n\nconst result = await workflowModeSkill.execute({\n action: 'detect',\n featureName: 'fix: minor bug'\n});\nconsole.log(result.detectedMode); // 'small'\n```\n\n---\n\n## π What's New in v5.8.0\n\n### CodeGraph MCP v0.8.0 Integration π\n\nMajor update to CodeGraph MCP integration with expanded language support and new features.\n\n#### 16 Language Support (NEW: Kotlin, Swift, Scala, Lua)\n\n- **Python, TypeScript, JavaScript** - Full AST analysis\n- **Rust, Go, Java, PHP, C#** - Enterprise language support\n- **C, C++, HCL (Terraform)** - System and infrastructure code\n- **Ruby** - Dynamic language support\n- **Kotlin** (NEW) - `.kt`, `.kts` files with classes, interfaces, functions\n- **Swift** (NEW) - `.swift` files with classes, structs, protocols\n- **Scala** (NEW) - `.scala`, `.sc` files with traits, objects\n- **Lua** (NEW) - `.lua` files with functions, table assignments\n\n#### File Watching & Auto Re-indexing\n\n- **`codegraph-mcp watch`** - Real-time file monitoring\n- **Debounce configuration** - Configurable delay (default: 1.0s)\n- **Community detection** - Optional `--community` flag after re-index\n\n#### Enhanced MCP Tools (14 Tools)\n\n| Category | Tools |\n|----------|-------|\n| **Graph Query** | `query_codebase`, `find_dependencies`, `find_callers`, `find_callees`, `find_implementations`, `analyze_module_structure` |\n| **Code Retrieval** | `get_code_snippet`, `read_file_content`, `get_file_structure` |\n| **GraphRAG** | `global_search`, `local_search` |\n| **Management** | `suggest_refactoring`, `reindex_repository`, `execute_shell_command` |\n\n#### Security & Performance (v0.7.3)\n\n- **Security fixes**: Path traversal and command injection protection\n- **Connection pooling**: Improved database performance\n- **Caching**: Faster repeated queries\n\n```javascript\n// Use CodeGraph MCP integration\nconst { CodeGraphIntegration } = require('musubi-sdd');\n\nconst cg = new CodeGraphIntegration('/path/to/repo');\nawait cg.indexRepository();\n\n// File watching (v0.7.0+)\nawait cg.startWatch({ debounce: 2.0, community: true });\n\n// Query codebase\nconst results = await cg.queryCodebase('authentication');\nconst callers = await cg.findCallers('UserService::login');\n```\n\n---\n\n## π What's New in v5.6.0\n\n### Enterprise-Scale Analysis & Rust Migration Support π’π¦\n\nMajor improvements based on analysis of GCC codebase (10+ million lines, 100,000+ files).\n\n#### Large Project Analyzer\n\n- **Scale-aware analysis**: Automatically detects project size and selects appropriate strategy\n- **Memory-efficient processing**: Chunk-based processing with garbage collection for 100K+ files\n- **Multi-language support**: JavaScript, TypeScript, C, C++, Python, Rust, Go, Java\n- **Giant function detection**: Flags functions with 100+ (warning), 500+ (critical), 1000+ (extreme) lines\n\n#### CodeGraph MCP Integration\n\n- **Deep code graph analysis**: Integration with CodeGraph MCP for relationship analysis\n- **Call graph generation**: Track callers and callees with configurable depth\n- **Impact analysis**: Identify affected files when code changes\n- **Hotspot identification**: Detect highly-connected entities (refactoring candidates)\n\n#### Enhanced Complexity Analyzer\n\n- **Cyclomatic complexity**: Standard decision-point counting\n- **Cognitive complexity**: SonarSource-style readability measurement\n- **Severity levels**: Ideal β Warning β Critical β Extreme thresholds\n\n#### Rust Migration Generator\n\n- **Unsafe pattern detection**: Identify C/C++ memory-unsafe patterns\n- **Migration priority scoring**: Automatic prioritization for Rust migration\n- **Security component analysis**: Flag security-critical code sections\n\n```javascript\n// Analyze enterprise-scale projects (100,000+ files)\nconst { LargeProjectAnalyzer, ComplexityAnalyzer } = require('musubi-sdd');\n\nconst analyzer = new LargeProjectAnalyzer('/path/to/gcc');\nconst result = await analyzer.analyze();\nconsole.log(result.stats); // { totalFiles: 109073, ... }\n\n// Calculate code complexity\nconst complexity = new ComplexityAnalyzer();\nconst score = complexity.calculateCyclomaticComplexity(code, 'javascript');\n```\n\n### v5.4.0: GitHub Reference & Repository Analysis π¦\n\n- π **`--reference` / `-r` Option** - Reference multiple GitHub repositories for pattern analysis\n- π **Repository Analysis** - Fetch metadata, structure, and key files via GitHub API\n- ποΈ **Pattern Detection** - Clean Architecture, Hexagonal, DDD, Monorepo patterns\n- π‘ **Improvement Suggestions** - Architecture and technology recommendations\n- π **Analysis Output** - Saved to `steering/references/github-references-YYYY-MM-DD.md`\n\n```bash\n# Reference multiple repositories for improvement suggestions\nmusubi init -r facebook/react -r vercel/next.js -r denoland/deno\n\n# With branch specification\nmusubi init --reference tokio-rs/tokio@main\n\n# Full URL format\nmusubi init --reference https://github.com/rust-lang/rust\n```\n\n### v5.3.0: Multi-Language Support π\n\n- π **Technology Stack Selection** - Single, Multiple, Undecided, or Help me decide\n- π€ **Language Recommendation Engine** - AI-powered suggestions based on app type, performance needs, team expertise\n- π **10 Languages Supported** - JS/TS, Python, Rust, Go, Java, C#, C++, Swift, Ruby, PHP\n- π **Dynamic tech.md** - Language-specific templates with frameworks and tools\n\n### v5.0.0: Phase 5 Advanced Features π\n\n- π **Steering Auto-Update** - Automatic project memory synchronization with 5 trigger types\n- β
**Steering Validator** - Comprehensive validation for steering documents\n- π **Template Constraints** - Enforce structure requirements with ThinkingChecklist\n- π **Quality Dashboard** - A-F grade metrics with health status monitoring\n- π **Advanced Validation** - Cross-artifact consistency, gap detection, traceability validation\n- π― **Phase 5 Integration** - Unified access to all advanced features\n\n### v4.0.0: Phase 4 Agent Loop & Agentic Features π€\n\n- π§ **Codebase Intelligence** - RepositoryMap, ASTExtractor, ContextOptimizer\n- π **Agentic Reasoning** - ReasoningEngine, PlanningEngine, SelfCorrection\n- β‘ **Code Generation** - CodeGenerator, CodeReviewer with multiple modes\n- π **Integrated Agent** - Unified agent with all Phase 4 capabilities\n\n```bash\n# Use advanced features via orchestrator\nconst { Phase5Integration, createIntegratedAgent } = require('musubi-sdd');\n\n# Quality dashboard\nconst dashboard = new QualityDashboard();\nconst metrics = await dashboard.collectMetrics(projectPath);\n\n# Advanced validation\nconst validator = new AdvancedValidation();\nconst result = await validator.validateAll(projectPath);\n```\n\n### Previous Highlights (v3.11.0)\n\n- π **WebSocket Real-time GUI** - Live replanning updates with `musubi-browser` dashboard\n- π **GUI Quick Actions** - Modal dialog for New Requirement, Validate Project, Export Report\n- π **GitHub Actions Integration** - `musubi-action` for CI/CD with MUSUBI validation\n- π§ **OpenAPI Converter** - Convert OpenAPI 3.x/Swagger 2.x specs to MUSUBI format\n- π **Multi-language Templates** - 7 language support (EN, JA, ZH, KO, ES, DE, FR)\n- π€ **Ollama Integration** - Local LLM support with 9 model presets\n\n### Previous (v3.6.0)\n\n- π§ **Dynamic Replanning Engine** - AI agents can now dynamically adjust plans when tasks fail\n- π **LLM Provider Abstraction** - Multi-provider support (Copilot, Anthropic, OpenAI)\n- π‘ **Real-time Plan Monitoring** - Detect failures, timeouts, and quality degradation\n- π **Alternative Path Generation** - LLM-powered alternative strategies with confidence scoring\n- π **Replan History & Audit** - Full audit trail with JSONL persistence and export\n\n### Previous (v3.5.1)\n\n- π **Workflow Engine** - New `musubi-workflow` CLI for stage management and metrics\n- π **Metrics Collection** - Track time per stage, iteration counts, feedback loops\n- π¬ **Spike/PoC Stage** - Stage 0 for research and prototyping before requirements\n- π **Code Review Stage** - Stage 5.5 between implementation and testing\n- π **Retrospective Stage** - Stage 9 for continuous improvement\n- β
**Stage Validation Guide** - Checklists for stage transition validation\n\n### Previous (v3.5.1)\n\n- π§ **CLI Integration** - Added CLI command references to all 8 Claude Code skills\n- π **Platform Documentation** - CLI Commands section added to all 6 non-Claude platforms\n\n### v2.1.0\n\n- π **CodeGraphMCPServer Integration** - 14 MCP tools for enhanced code analysis\n- π§ **GraphRAG-Powered Search** - Semantic code understanding with Louvain community detection\n- π **11 Agents Enhanced** - Key agents now leverage MCP tools for deeper code analysis\n\n## Features\n\n- π€ **Multi-Agent Support** - Works with 7 AI coding agents (Claude Code, GitHub Copilot, Cursor, Gemini CLI, Codex CLI, Qwen Code, Windsurf)\n- π§ **Dynamic Replanning** - AI agents dynamically adjust plans on failure with LLM-powered alternatives (v3.6.0+)\n- π **MCP Server Integration** - CodeGraphMCPServer for advanced code analysis (v2.0.0)\n- π **Flexible Command Formats** - Supports Markdown, TOML, and AGENTS.md formats\n- π― **27 Specialized Skills (All Platforms)** - 25 platform agents + 5 orchestrator built-in skills (v5.9.0)\n - Claude Code: Skills API (25 skills + 5 built-in)\n - GitHub Copilot & Cursor: AGENTS.md (official support)\n - Other 4 agents: AGENTS.md (compatible format)\n- π **Constitutional Governance** - 9 immutable articles + Phase -1 Gates for quality enforcement\n- π **EARS Requirements Generator** - Create unambiguous requirements with 5 EARS patterns (v0.8.0)\n- ποΈ **Design Document Generator** - Create C4 models and ADRs with traceability (v0.8.2)\n- π **Change Management System** - Delta specifications for brownfield projects (v0.8.6)\n- π **Gap Detection System** - Identify orphaned requirements and untested code (v0.8.7)\n- π§ **Auto-Updating Project Memory** - Steering system maintains architecture, tech stack, and product context\n- π **Automatic Onboarding** - `musubi-onboard` analyzes existing projects and generates steering docs (2-5 minutes)\n- π **Auto-Sync** - `musubi-sync` detects codebase changes and keeps steering docs current\n- π **Intelligent Code Analysis** - `musubi-analyze` provides quality metrics, complexity analysis, and technical debt detection\n- π€ **Team Collaboration** - `musubi-share` enables memory sharing, import/export, and multi-platform sync (v0.6.0)\n- β
**Constitutional Validation** - `musubi-validate` enforces 9 immutable governance articles with Phase -1 Gates (v0.7.0)\n- β
**Complete Traceability** - Requirements β Design β Code β Tests mapping\n- π **Bilingual Documentation** - All agent-generated documents created in both English and Japanese\n\n## Supported AI Coding Agents\n\nMUSUBI supports 7 AI coding agents, each with tailored configurations:\n\n| Agent | Skills API | 27 Skills | Command Format | Command File Format | Installation Directory |\n| ------------------ | -------------- | -------------- | ---------------- | ------------------------- | --------------------------------------------- |\n| **Claude Code** | β
(27 skills) | β
| `/sdd-*` | Markdown (.md) | `.claude/skills/`, `.claude/commands/` |\n| **GitHub Copilot** | β | β
(AGENTS.md) | `/sdd-*` | Prompt File (.prompt.md) | `.github/prompts/`, `.github/AGENTS.md` |\n| **Cursor IDE** | β | β
(AGENTS.md) | `/sdd-*` | Markdown + AGENTS.md | `.cursor/commands/`, `.cursor/AGENTS.md` |\n| **Gemini CLI** | β | β
(GEMINI.md) | `/sdd-*` | TOML + GEMINI.md | `.gemini/commands/`, `GEMINI.md` |\n| **Codex CLI** | β | β
(AGENTS.md) | `/prompts:sdd-*` | Markdown + AGENTS.md | `.codex/prompts/`, `.codex/AGENTS.md` |\n| **Qwen Code** | β | β
(AGENTS.md) | `/sdd-*` | Markdown + AGENTS.md | `.qwen/commands/`, `.qwen/AGENTS.md` |\n| **Windsurf IDE** | β | β
(AGENTS.md) | `/sdd-*` | Markdown + AGENTS.md | `.windsurf/workflows/`, `.windsurf/AGENTS.md` |\n\n**Notes**:\n\n- Skills API is exclusive to Claude Code\n- **All 7 platforms now support 27 skills** via Skills API (Claude Code) or AGENTS.md (others)\n- v5.9.0 added 5 built-in orchestrator skills (release, workflow, package, constitution-level, project-config)\n- **GitHub Copilot uses `.prompt.md` extension** per VS Code official documentation\n- AGENTS.md: OpenAI specification, officially supported by GitHub Copilot & Cursor\n- Gemini CLI uses TOML format + GEMINI.md integration\n\n---\n\n## Constitutional Governance\n\nMUSUBI enforces 9 Constitutional Articles for quality assurance:\n\n```bash\n# Validate constitutional compliance\nmusubi-validate all\nmusubi-validate constitution\nmusubi-validate gates\nmusubi-validate complexity\n```\n\n**9 Articles**:\n\n1. **Library-First Principle** - All features begin as independent libraries\n2. **CLI Interface Mandate** - All libraries expose CLI functionality\n3. **Test-First Imperative** - Tests written before code (80% coverage required)\n4. **EARS Requirements Format** - 5 EARS patterns for unambiguous requirements\n5. **Traceability Mandate** - 100% traceability: Requirements β Design β Code β Tests\n6. **Project Memory** - Steering system maintains project context\n7. **Simplicity Gate** - Maximum 3 sub-projects initially (Phase -1 Gate)\n8. **Anti-Abstraction Gate** - Use framework APIs directly (Phase -1 Gate)\n9. **Integration-First Testing** - Integration tests use real services (no mocks)\n\n**Phase -1 Gates**: Pre-implementation validation checkpoints for Articles VII & VIII. See:\n\n- [steering/rules/constitution.md](steering/rules/constitution.md) - Full constitutional text\n- [steering/rules/phase-gates.md](steering/rules/phase-gates.md) - Approval process and active gates\n- All other agents use Markdown format + AGENTS.md\n\n## Quick Start\n\n### Installation via npx\n\n```bash\n# Initialize MUSUBI for your preferred agent\n\n# Claude Code (default) - 27 Skills (25 + 5 built-in)\nnpx musubi-sdd init\nnpx musubi-sdd init --claude\n\n# GitHub Copilot - 27 skills (AGENTS.md, official support)\nnpx musubi-sdd init --copilot\n\n# Cursor IDE - 27 skills (AGENTS.md, official support)\nnpx musubi-sdd init --cursor\n\n# Gemini CLI - 27 skills (GEMINI.md integration)\nnpx musubi-sdd init --gemini\n\n# Codex CLI - 27 skills (AGENTS.md)\nnpx musubi-sdd init --codex\n\n# Qwen Code - 27 skills (AGENTS.md)\nnpx musubi-sdd init --qwen\n\n# Windsurf IDE - 27 skills (AGENTS.md)\nnpx musubi-sdd init --windsurf\n\n# Or install globally\nnpm install -g musubi-sdd\nmusubi init --claude # or --copilot, --cursor, etc.\n\n# Onboard existing project (automatic analysis)\nmusubi-onboard\n\n# Synchronize steering docs with codebase\nmusubi-sync\nmusubi-sync --dry-run # Preview changes\nmusubi-sync --auto-approve # Auto-apply (CI/CD)\n\n# Analyze code quality (v0.5.0)\nmusubi-analyze # Full analysis\nmusubi-analyze --type=quality # Quality metrics only\nmusubi-analyze --type=dependencies # Dependencies only\nmusubi-analyze --type=security # Security audit\nmusubi-analyze --output=report.md # Save report\n\n# Share project memories with team (v0.6.0)\nmusubi-share export # Export memories to JSON\nmusubi-share import memories.json # Import from teammate\nmusubi-share sync --platform=copilot # Sync to specific platform\n\n# Validate constitutional compliance (v0.7.0)\nmusubi-validate constitution # Validate all 9 articles\nmusubi-validate article 3 # Validate Test-First Imperative\nmusubi-validate gates # Validate Phase -1 Gates\nmusubi-validate complexity # Check complexity limits\nmusubi-validate all -v # Full validation with details\n\n# Generate EARS requirements documents (v0.8.0)\nmusubi-requirements init \"User Authentication\" # Initialize requirements doc\nmusubi-requirements add # Add requirement interactively\nmusubi-requirements list # List all requirements\nmusubi-requirements validate # Validate EARS format\nmusubi-requirements trace # Show traceability matrix\n\n# Generate design documents (v0.8.2)\nmusubi-design init \"User Authentication\" # Initialize design document\nmusubi-design add-c4 context # Add C4 Context diagram\nmusubi-design add-c4 container --format plantuml # Add Container with PlantUML\nmusubi-design add-adr \"Use JWT for tokens\" # Add Architecture Decision\nmusubi-design validate # Validate design completeness\nmusubi-design trace # Show requirements traceability\n\n# Break down design into tasks (v0.8.4)\nmusubi-tasks init \"User Authentication\" # Initialize task breakdown\nmusubi-tasks add \"Database Schema\" # Add task interactively\nmusubi-tasks list # List all tasks\nmusubi-tasks list --priority P0 # List critical tasks\nmusubi-tasks update 001 \"In Progress\" # Update task status\nmusubi-tasks validate # Validate task completeness\nmusubi-tasks graph # Show dependency graph\n\n# End-to-end traceability (v0.8.5)\nmusubi-trace matrix # Generate traceability matrix\nmusubi-trace matrix --format markdown > trace.md # Export to markdown\nmusubi-trace coverage # Calculate coverage statistics\nmusubi-trace coverage --min-coverage 100 # Require 100% coverage\nmusubi-trace gaps # Detect orphaned requirements/code\nmusubi-trace requirement REQ-AUTH-001 # Trace specific requirement\nmusubi-trace validate # Validate 100% traceability (Article V)\nmusubi-trace bidirectional # Bidirectional traceability analysis (v0.9.4)\nmusubi-trace impact REQ-AUTH-001 # Impact analysis for requirement changes (v0.9.4)\nmusubi-trace statistics # Comprehensive project statistics (v0.9.4)\n\n# Change management for brownfield projects (v0.8.6)\nmusubi-change init CHANGE-001 --title \"Add authentication\" # Create change proposal\nmusubi-change validate CHANGE-001 --verbose # Validate delta specification\nmusubi-change apply CHANGE-001 --dry-run # Preview changes\nmusubi-change apply CHANGE-001 # Apply changes to codebase\nmusubi-change archive CHANGE-001 # Archive to specs/\nmusubi-change list --status pending # List pending changes\nmusubi-change list --format json # List in JSON format\n\n# Gap detection and coverage validation (v0.8.7)\nmusubi-gaps detect # Detect all gaps\nmusubi-gaps detect --verbose # Show detailed gap information\nmusubi-gaps requirements # Detect orphaned requirements\nmusubi-gaps code # Detect untested code\nmusubi-gaps coverage # Calculate coverage statistics\nmusubi-gaps coverage --min-coverage 100 # Require 100% coverage\nmusubi-gaps detect --format markdown > gaps.md # Export gap report\n```\n\n## Documentation\n\nComprehensive guides are available in `docs/guides/`:\n\n- **[Brownfield Tutorial](docs/guides/brownfield-tutorial.md)** - Step-by-step guide for managing changes in existing projects\n- **[Delta Specification Guide](docs/guides/delta-spec-guide.md)** - Format reference for change tracking\n- **[Change Management Workflow](docs/guides/change-management-workflow.md)** - End-to-end workflow documentation\n- **[Traceability Matrix Guide](docs/guides/traceability-matrix-guide.md)** - Traceability system usage\n- **[Video Tutorial Plan](docs/guides/video-tutorial-plan.md)** - Video content script\n\n### Project Types\n\nDuring initialization, MUSUBI asks you to select a **Project Type**. This determines the workflow and features available:\n\n#### Greenfield (0β1)\n\n- **What it is**: Starting a new project from scratch\n- **Use cases**:\n - New application development\n - Proof-of-concept projects\n - Greenfield microservices\n- **Features enabled**:\n - Full 8-stage SDD workflow (Research β Monitoring)\n - `/sdd-steering` - Generate initial project memory\n - `/sdd-requirements` - Create new requirements from scratch\n - `/sdd-design` - Design architecture (C4 model + ADR)\n - `/sdd-tasks` - Break requirements into tasks\n - `/sdd-implement` - Implement features (test-first)\n - `/sdd-validate` - Constitutional compliance checks\n- **Benefits**:\n - Clean start with best practices enforced\n - Constitutional governance from day one\n - Complete traceability from requirements to code\n\n#### Brownfield (1βn)\n\n- **What it is**: Working with existing codebases\n- **Use cases**:\n - Adding features to existing applications\n - Refactoring legacy code\n - Migrating/modernizing systems\n- **Features enabled**:\n - Delta Specifications (ADDED/MODIFIED/REMOVED)\n - `/sdd-change-init` - Create change proposal\n - `/sdd-change-apply` - Apply changes with impact analysis\n - `/sdd-change-archive` - Archive completed changes\n - `change-impact-analyzer` skill (Claude Code) - Automatic impact detection\n - Reverse engineering: `/sdd-steering` analyzes existing code\n- **Benefits**:\n - Safe incremental changes with impact analysis\n - Preserves existing architecture while improving incrementally\n - Full audit trail of what changed and why\n\n#### Both\n\n- **What it is**: Hybrid approach for complex scenarios\n- **Use cases**:\n - Monolith β Microservices migration (brownfield + greenfield services)\n - Platform modernization (keep some, rebuild others)\n - Multi-component systems with mixed maturity\n- **Features enabled**:\n - All Greenfield + Brownfield features\n - Flexibility to choose workflow per component\n - Mixed delta specs and greenfield specs in same project\n- **Benefits**:\n - Maximum flexibility for complex transformation projects\n - Unified steering/governance across all components\n - Single tool for entire modernization journey\n\n**Example Selection**:\n\n```text\n? Project type:\nβ― Greenfield (0β1) β New projects\n Brownfield (1βn) β Existing codebases\n Both β Complex/hybrid scenarios\n```\n\n### What Gets Installed\n\n#### Claude Code (Skills API)\n\n```text\nyour-project/\nβββ .claude/\nβ βββ skills/ # 25 Skills API + 5 built-in (Claude Code exclusive)\nβ β βββ orchestrator/\nβ β βββ steering/\nβ β βββ requirements-analyst/\nβ β βββ ... (22 more)\nβ βββ commands/ # Slash commands (/sdd-*)\nβ βββ CLAUDE.md # Claude Code guide\nβββ steering/ # Project memory (all agents)\nβ βββ project.yml # Project configuration (v0.2.1+)\nβ βββ structure.md # Architecture patterns\nβ βββ tech.md # Technology stack\nβ βββ product.md # Product context\nβ βββ memories/ # Persistent knowledge (v0.2.0+)\nβ β βββ architecture_decisions.md\nβ β βββ development_workflow.md\nβ β βββ domain_knowledge.md\nβ β βββ lessons_learned.md\nβ β βββ suggested_commands.md\nβ β βββ technical_debt.md\nβ βββ rules/\nβ βββ constitution.md # 9 Constitutional Articles\nβ βββ workflow.md # 8-Stage SDD workflow\nβ βββ ears-format.md # EARS syntax guide\nβββ templates/ # Document templates (all agents)\nβββ storage/ # Specs, changes, features (all agents)\n```\n\n#### Other Agents (GitHub Copilot, Cursor, Gemini, etc.)\n\n```text\nyour-project/\nβββ .github/prompts/ # For GitHub Copilot (/sdd-*, .prompt.md)\nβ βββ sdd-steering.prompt.md # Prompt files use .prompt.md extension\nβ βββ sdd-requirements.prompt.md\nβ βββ ... (9 prompt files)\nβ βββ AGENTS.md # 27 skills definition (official support)\nβ OR\nβββ .cursor/commands/ # For Cursor (/sdd-*, Markdown)\nβ βββ AGENTS.md # 27 skills definition (official support)\nβ OR\nβββ .gemini/commands/ # For Gemini CLI (/sdd-*, TOML)\nβ β βββ sdd-steering.toml\nβ β βββ sdd-requirements.toml\nβ β βββ ... (6 TOML files)\nβ OR\nβββ .codex/prompts/ # For Codex CLI (/prompts:sdd-*, Markdown)\nβ βββ AGENTS.md # 27 skills definition\nβ OR\nβββ .qwen/commands/ # For Qwen Code (/sdd-*, Markdown)\nβ βββ AGENTS.md # 27 skills definition\nβ OR\nβββ .windsurf/workflows/ # For Windsurf (/sdd-*, Markdown)\nβ βββ AGENTS.md # 27 skills definition\nβ\nβββ GEMINI.md (root, for Gemini) # 27 skills integrated into existing file\nβββ steering/ # Project memory (same for all)\nβ βββ project.yml # Project configuration (v0.2.1+)\nβ βββ memories/ # Persistent knowledge (v0.2.0+)\nβ β βββ architecture_decisions.md\nβ β βββ development_workflow.md\nβ β βββ domain_knowledge.md\nβ β βββ lessons_learned.md\nβ β βββ suggested_commands.md\nβ β βββ technical_debt.md\nβ βββ ... (structure.md, tech.md, product.md, rules/)\nβββ templates/ # Document templates (same for all)\nβββ storage/ # Specs, changes, features (same for all)\n```\n\n**Key Differences**:\n\n- **Claude Code**: 27 Skills (25 + 5 built-in) + commands (Markdown)\n- **GitHub Copilot**: AGENTS.md (official support) + prompt files (`.prompt.md`)\n- **Cursor**: AGENTS.md (official support) + commands (Markdown)\n- **Gemini CLI**: GEMINI.md integration (27 skills) + TOML commands (unique)\n- **Others**: AGENTS.md (compatible) + Markdown commands\n- **All platforms**: Same 27 skills, different implementation formats\n\n## Usage\n\n### CLI Commands\n\nMUSUBI provides several CLI commands for project management:\n\n```bash\n# Show version\nmusubi --version\nmusubi -v\n\n# Show help\nmusubi --help\n\n# Show comprehensive info\nmusubi info\n\n# Check project status\nmusubi status\n\n# Validate constitutional compliance\nmusubi validate\nmusubi validate --verbose # Detailed output\nmusubi validate --all # Validate all features\n\n# Initialize MUSUBI (interactive)\nmusubi init\n\n# Onboard existing project (v0.3.0+)\nmusubi-onboard\nmusubi-onboard --auto-approve # Skip confirmation\nmusubi-onboard --skip-memories # Skip memory initialization\n\n# Synchronize steering docs with codebase (v0.4.0+)\nmusubi-sync # Interactive mode\nmusubi-sync --dry-run # Preview changes only\nmusubi-sync --auto-approve # Auto-apply (CI/CD)\n\n# Analyze code quality (v0.5.0+)\nmusubi-analyze # Full analysis (quality + dependencies + security)\nmusubi-analyze --type=quality # Code quality metrics only\nmusubi-analyze --type=dependencies # Dependency analysis\nmusubi-analyze --type=security # Security vulnerabilities\nmusubi-analyze --output=report.md # Save report to file\nmusubi-analyze --json # JSON output\n\n# Share project memories with team (v0.6.0+)\nmusubi-share export # Export memories to JSON/YAML\nmusubi-share export --output=memories.yaml # YAML format\nmusubi-share import memories.json # Import and merge\nmusubi-share import memories.json --strategy=theirs # Auto-accept\nmusubi-share sync # Sync across AI platforms\nmusubi-share sync --platform=cursor # Sync specific platform\nmusubi-share status # Show sharing status\n```\n\n#### musubi-onboard\n\nAutomatically analyzes existing projects and generates steering documentation:\n\n```text\nπ MUSUBI Onboarding Wizard\n\nAnalyzing your project...\n\nβ
Project structure analyzed\nβ
Technology stack detected\n - Node.js, TypeScript, React, Jest\nβ
Steering documents generated\n - steering/structure.md (en + ja)\n - steering/tech.md (en + ja)\n - steering/product.md (en + ja)\nβ
Memories initialized (6 files)\nβ
Project configuration created\n - steering/project.yml\n\nβ±οΈ Onboarding completed in 2.5 minutes\n\nπ‘ Next steps:\n - Review generated steering docs\n - Run: musubi-sync to keep docs current\n - Create requirements: /sdd-requirements [feature]\n```\n\n**Features**:\n\n- Automatic codebase analysis (package.json, directory structure)\n- Technology stack detection (languages, frameworks)\n- Bilingual steering docs generation (English + Japanese)\n- Memory system initialization (6 memory files)\n- Project configuration (project.yml)\n- 96% time reduction (2-4 hours β 2-5 minutes)\n\n#### musubi-sync\n\nDetects codebase changes and keeps steering documents synchronized:\n\n```text\nπ MUSUBI Steering Sync\n\nDetected changes:\n π¦ Version: 0.3.0 β 0.4.0\n β New framework: js-yaml@4.1.0\n π New directory: bin/\n\n? Apply these changes? (Y/n) Y\n\nβ
Updated steering/project.yml\nβ
Updated steering/tech.md (en + ja)\nβ
Updated steering/structure.md (en + ja)\nβ
Recorded change in memories/architecture_decisions.md\n\nπ Steering synchronized successfully!\n```\n\n**Features**:\n\n- Change detection (version, languages, frameworks, directories)\n- Interactive mode (default): Show changes, ask confirmation\n- Auto-approve mode (--auto-approve): CI/CD integration\n- Dry-run mode (--dry-run): Preview only\n- Bilingual updates (English + Japanese together)\n- Memory recording (audit trail)\n\n**Usage**:\n\n```bash\n# Interactive (default)\nmusubi-sync\n\n# Preview changes without applying\nmusubi-sync --dry-run\n\n# Auto-apply for CI/CD pipelines\nmusubi-sync --auto-approve\n```\n\n#### musubi-share\n\nShare and merge project memories across team members and AI platforms (v0.6.0+):\n\n```text\nπ€ MUSUBI Memory Export\n\nExport Summary:\n File: team-memories.json\n Format: json\n Size: 1098.28 KB\n Memories: 7 files\n Agents: 1 platforms\n```\n\n**Features**:\n\n- **Export**: Share memories as JSON/YAML\n- **Import**: Merge memories from teammates\n- **Sync**: Synchronize across AI platforms\n- **Conflict Resolution**: Interactive, auto-accept, keep-local, or merge with markers\n- **Status**: Show installed platforms and memory counts\n\n**Usage**:\n\n```bash\n# Export memories\nmusubi-share export\nmusubi-share export --output=memories.yaml\n\n# Import and merge\nmusubi-share import colleague-memories.json\nmusubi-share import memories.json --strategy=theirs # Auto-accept\nmusubi-share import memories.json --strategy=ours # Keep local\nmusubi-share import memories.json --strategy=merge # With markers\n\n# Platform sync\nmusubi-share sync\nmusubi-share sync --platform=cursor\n\n# Status check\nmusubi-share status\n```\n\n#### musubi status\n\nShows the current state of your MUSUBI project:\n\n```text\nπ MUSUBI Project Status\n\nβ
MUSUBI is initialized\n\nπ Claude Code Skills: 27 installed (25 + 5 built-in)\n Location: .claude/skills/\n\nπ§ Steering Context:\n β
structure.md (updated: 2025-11-16)\n β
tech.md (updated: 2025-11-16)\n β
product.md (updated: 2025-11-16)\n\nβ
Constitutional Governance: Enabled\n\nπ Specifications: 3 documents\n Latest specs:\n - auth-requirements.md\n - auth-design.md\n - auth-tasks.md\n\nπ‘ Next steps:\n - Review steering files in steering/\n - Create requirements: /sdd-requirements [feature]\n - Validate compliance: musubi validate\n```\n\n#### musubi validate\n\nPerforms quick constitutional compliance checks:\n\n- **Article I**: Library-First Principle (checks `lib/` directory)\n- **Article II**: CLI Interface Mandate (checks for `cli.ts` files)\n- **Article IV**: EARS Requirements Format (validates EARS patterns)\n- **Article VI**: Project Memory (checks steering files)\n\nFor comprehensive validation, use your agent's `/sdd-validate` (or equivalent) command.\n\n### Agent-Specific Commands\n\n#### Claude Code\n\n```bash\n# Generate project memory\n/sdd-steering\n\n# Create requirements\n/sdd-requirements authentication\n\n# Design architecture\n/sdd-design authentication\n\n# Break down into tasks\n/sdd-tasks authentication\n\n# Implement feature\n/sdd-implement authentication\n\n# Validate constitutional compliance\n/sdd-validate authentication\n```\n\n**Skills (Auto-Invoked)**: Claude Code automatically selects the appropriate skill:\n\n- \"Review my code\" β `code-reviewer` skill\n- \"Create requirements for user login\" β `requirements-analyst` skill\n- \"Design API for payment\" β `api-designer` skill\n\n#### GitHub Copilot\n\n```bash\n# Use # prefix for custom prompts\n#sdd-steering\n#sdd-requirements authentication\n#sdd-design authentication\n#sdd-tasks authentication\n#sdd-implement authentication\n#sdd-validate authentication\n```\n\n#### Gemini CLI\n\n```bash\n# Use / prefix for commands (TOML format)\n/sdd-steering\n/sdd-requirements authentication\n/sdd-design authentication\n/sdd-tasks authentication\n/sdd-implement authentication\n/sdd-validate authentication\n```\n\n**Note**: Gemini CLI commands are defined in TOML format (`.toml` files) instead of Markdown.\n\n#### Cursor IDE, Qwen Code, Windsurf\n\n```bash\n# Use / prefix for commands (Markdown format)\n/sdd-steering\n/sdd-requirements authentication\n/sdd-design authentication\n/sdd-tasks authentication\n/sdd-implement authentication\n/sdd-validate authentication\n```\n\n#### Codex CLI\n\n```bash\n# Use /prompts: prefix\n/prompts:sdd-steering\n/prompts:sdd-requirements authentication\n/prompts:sdd-design authentication\n/prompts:sdd-tasks authentication\n/prompts:sdd-implement authentication\n/prompts:sdd-validate authentication\n```\n\n## 27 Skills Overview (All Platforms)\n\n**Available on all 7 platforms** via:\n\n- **Claude Code**: Skills API (automatic invocation) + 5 built-in orchestrator skills\n- **GitHub Copilot & Cursor**: AGENTS.md (official support, reference via `@agent-name`)\n- **Gemini, Windsurf, Codex, Qwen**: AGENTS.md (compatible format, natural language reference)\n\n### Orchestration & Management (3 + 5 built-in)\n\n- **orchestrator** - Master coordinator for multi-skill workflows\n- **steering** - Project memory manager (auto-updating context)\n- **constitution-enforcer** - Governance validation (9 Articles + Phase -1 Gates)\n\n#### Built-in Orchestrator Skills (v5.9.0)\n\n- **release-manager** - CHANGELOG generation, version management\n- **workflow-mode-manager** - Mode detection (small/medium/large)\n- **package-manager** - Monorepo package & dependency analysis\n- **constitution-level-manager** - Level-aware validation (critical/advisory/flexible)\n- **project-config-manager** - Schema validation & migration\n\n### Requirements & Planning (3)\n\n- **requirements-analyst** - EARS format requirements generation\n- **project-manager** - Project planning, scheduling, risk management\n- **change-impact-analyzer** - Brownfield change analysis\n\n### Architecture & Design (4)\n\n- **system-architect** - C4 model + ADR architecture design\n- **api-designer** - REST/GraphQL/gRPC API design\n- **database-schema-designer** - Database design, ER diagrams, DDL\n- **ui-ux-designer** - UI/UX design, wireframes, prototypes\n\n### Development (1)\n\n- **software-developer** - Multi-language code implementation\n\n### Quality & Review (5)\n\n- **test-engineer** - Unit, integration, E2E testing with EARS mapping\n- **code-reviewer** - Code review, SOLID principles\n- **bug-hunter** - Bug investigation, root cause analysis\n- **quality-assurance** - QA strategy, test planning\n- **traceability-auditor** - Requirements β Code β Test coverage validation\n\n### Security & Performance (2)\n\n- **security-auditor** - OWASP Top 10, vulnerability detection\n- **performance-optimizer** - Performance analysis, optimization\n\n### Infrastructure & Operations (5)\n\n- **devops-engineer** - CI/CD pipelines, Docker/Kubernetes\n- **cloud-architect** - AWS/Azure/GCP, IaC (Terraform/Bicep)\n- **database-administrator** - Database operations, tuning\n- **site-reliability-engineer** - Production monitoring, SLO/SLI, incident response\n- **release-coordinator** - Multi-component release management\n\n### Documentation & Specialized (2)\n\n- **technical-writer** - Technical documentation, API docs\n- **ai-ml-engineer** - ML model development, MLOps\n\n## Governance Articles\n\nMUSUBI enforces 9 immutable constitutional articles:\n\n1. **Library-First Principle** - Features start as libraries\n2. **CLI Interface Mandate** - All libraries expose CLI\n3. **Test-First Imperative** - Tests before code (Red-Green-Blue)\n4. **EARS Requirements Format** - Unambiguous requirements\n5. **Traceability Mandate** - 100% coverage required\n6. **Project Memory** - All skills check steering first\n7. **Simplicity Gate** - Maximum 3 projects initially\n8. **Anti-Abstraction Gate** - Use framework features directly\n9. **Integration-First Testing** - Real services over mocks\n\n## SDD Workflow (8 Stages)\n\n```text\n1. Research β 2. Requirements β 3. Design β 4. Tasks β\n5. Implementation β 6. Testing β 7. Deployment β 8. Monitoring\n```\n\nEach stage has:\n\n- Dedicated skills\n- Quality gates\n- Traceability requirements\n- Constitutional validation\n\n## EARS Requirements Format\n\n```markdown\n### Requirement: User Login\n\nWHEN user provides valid credentials,\nTHEN the system SHALL authenticate the user\nAND the system SHALL create a session.\n\n#### Scenario: Successful login\n\n- WHEN user enters correct email and password\n- THEN system SHALL verify credentials\n- AND system SHALL redirect to dashboard\n```\n\n## Bilingual Documentation\n\n**All agent-generated documents are created in both English and Japanese.**\n\n### Language Policy\n\n- **English**: Reference/source documents (`.md`)\n- **Japanese**: Translations (`.ja.md`)\n- **Skills**: Always read English versions for work\n- **Code References**: Requirement IDs, technical terms stay in English\n\n### Files Generated Bilingually\n\n**Steering Context**:\n\n- `steering/structure.md` + `steering/structure.ja.md`\n- `steering/tech.md` + `steering/tech.ja.md`\n- `steering/product.md` + `steering/product.ja.md`\n\n**Specifications**:\n\n- `storage/specs/auth-requirements.md` + `storage/specs/auth-requirements.ja.md`\n- `storage/design/auth-design.md` + `storage/design/auth-design.ja.md`\n- `storage/tasks/auth-tasks.md` + `storage/tasks/auth-tasks.ja.md`\n\n### Generation Order\n\n1. **English version generated FIRST** (reference/source)\n2. **Japanese version generated SECOND** (translation)\n3. Technical terms (REQ-XXX-NNN, EARS keywords, API endpoints) remain in English\n4. Both versions maintained in sync\n\n## Delta Specifications (Brownfield)\n\n```markdown\n## ADDED Requirements\n\n### REQ-NEW-001: Two-Factor Authentication\n\n...\n\n## MODIFIED Requirements\n\n### REQ-001: User Authentication\n\n**Previous**: Email + password\n**Updated**: Email + password + OTP\n...\n\n## REMOVED Requirements\n\n### REQ-OLD-005: Remember Me\n\n**Reason**: Security policy change\n```\n\n## Example Usage\n\n### Greenfield Project (0β1)\n\n```bash\n# 1. Initialize\nnpx musubi-sdd init\n\n# 2. Generate steering\n/sdd-steering\n\n# 3. Create requirements\n/sdd-requirements user-authentication\n\n# 4. Design architecture\n/sdd-design user-authentication\n\n# 5. Break into tasks\n/sdd-tasks user-authentication\n\n# 6. Implement\n/sdd-implement user-authentication\n```\n\n### Brownfield Project (1βn)\n\n```bash\n# 1. Initialize with existing codebase\nnpx musubi-sdd init\n\n# 2. Generate steering from existing code\n/sdd-steering\n\n# 3. Create change proposal\n/sdd-change-init add-2fa\n\n# 4. Impact analysis (automatic via change-impact-analyzer skill)\n\n# 5. Implement change\n/sdd-change-apply add-2fa\n\n# 6. Archive change\n/sdd-change-archive add-2fa\n```\n\n## Configuration\n\n### MCP Server Integration\n\nMUSUBI v2.0.0 integrates with **CodeGraphMCPServer** for advanced code analysis.\n\n#### Option 1: Claude Code (Terminal)\n\n```bash\n# Install CodeGraph MCP with pipx (--force ensures latest version)\npipx install --force codegraph-mcp-server\n\n# Add to Claude Code\nclaude mcp add codegraph -- codegraph-mcp serve --repo .\n\n# Verify installation\nclaude mcp list\n```\n\n#### Option 2: VS Code with Claude Extension\n\n1. **Install Prerequisites**:\n\n ```bash\n # --force ensures latest version even if already installed\n pipx install --force codegraph-mcp-server\n ```\n\n2. **Configure VS Code** (`.vscode/mcp.json`):\n\n ```json\n {\n \"servers\": {\n \"codegraph\": {\n \"type\": \"stdio\",\n \"command\": \"codegraph-mcp\",\n \"args\": [\"serve\", \"--repo\", \"${workspaceFolder}\"]\n }\n }\n }\n ```\n\n3. **Or use Claude Desktop config** (`~/.claude/claude_desktop_config.json` on macOS/Linux, `%APPDATA%\\Claude\\claude_desktop_config.json` on Windows):\n\n ```json\n {\n \"mcpServers\": {\n \"codegraph\": {\n \"command\": \"codegraph-mcp\",\n \"args\": [\"serve\", \"--repo\", \"/path/to/your/project\"]\n }\n }\n }\n ```\n\n#### Option 3: npx (No Installation)\n\n```bash\n# Add via npx (no global install needed)\nclaude mcp add codegraph -- npx -y @anthropic/codegraph-mcp --codebase .\n```\n\n#### Verify MCP Server is Working\n\nAfter setup, test in Claude:\n\n```text\nUse the init_graph tool to analyze this codebase\n```\n\nIf successful, you'll see the code graph initialization output.\n\n**Available MCP Tools (14 tools)**:\n\n| Category | Tools | Description |\n| ---------- | --------------------------------------------------------------------- | -------------------------------- |\n| Code Graph | `init_graph`, `get_code_snippet`, `find_callers`, `find_dependencies` | Build and query code graph |\n| Search | `local_search`, `global_search`, `query_codebase` | GraphRAG-powered semantic search |\n| Analysis | `analyze_module_structure`, `suggest_refactoring` | Code structure analysis |\n| Navigation | `jump_to_definition`, `find_implementations` | Code navigation |\n\n**Agent Γ MCP Tool Mapping**:\n\n| Agent | Primary MCP Tools | Use Case |\n| ----------------------- | ------------------------------------------- | -------------------------------- |\n| @change-impact-analyzer | `find_dependencies`, `find_callers` | Impact analysis |\n| @traceability-auditor | `query_codebase`, `find_callers` | Traceability validation |\n| @system-architect | `analyze_module_structure`, `global_search` | Architecture analysis |\n| @code-reviewer | `suggest_refactoring`, `get_code_snippet` | Code quality review |\n| @security-auditor | `find_callers`, `query_codebase` | Security vulnerability detection |\n\nAlso integrates with other MCP servers:\n\n- **Context7 MCP** - Up-to-date library documentation (Next.js, React, etc.)\n- **Azure MCP** - Azure resource management\n- **Microsoft Learn MCP** - Microsoft documentation\n\nSkills automatically use available MCP servers when needed.\n\n### Customization\n\nEdit steering files to customize for your project:\n\n```bash\n# Architecture patterns\nsteering/structure.md\n\n# Technology stack\nsteering/tech.md\n\n# Product context\nsteering/product.md\n\n# Constitutional rules (if needed)\nsteering/rules/constitution.md\n```\n\n## Development\n\n### Contributing Workflow\n\n1. Fork the repository\n2. Create a feature branch: `git checkout -b feature/your-feature`\n3. Make your changes\n4. Run tests locally: `npm test`\n5. Run lint: `npm run lint`\n6. Commit with conventional commits: `git commit -m \"feat: add new feature\"`\n7. Push to your fork: `git push origin feature/your-feature`\n8. Create a Pull Request\n9. Wait for CI checks to pass (all checks must succeed)\n10. Request review\n11. Merge after approval\n\n### CI/CD Pipeline\n\n- **CI**: Runs on every PR and push to `main`\n - ESLint & Prettier\n - Jest Tests (80% coverage required)\n - Build Verification\n - Security Audit\n - Platform Initialization Tests (7 platforms)\n- **Release**: Automated npm publish on version tags (`v*.*.*`)\n- **Dependabot**: Weekly dependency updates (Mondays 9:00 JST)\n\n### Local Testing\n\n```bash\n# Clone repository\ngit clone https://github.com/nahisaho/MUSUBI.git\ncd musubi\n\n# Install dependencies\nnpm install\n\n# Run tests\nnpm test\n\n# Run lint\nnpm run lint\n\n# Check formatting\nnpm run format:check\n\n# Link for local development\nnpm link\nmusubi init\n```\n\n## Contributing\n\nContributions welcome! See [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines.\n\n## β Support\n\nIf you find MUSUBI useful, please consider:\n\n- β **Star this repository** - It helps others discover MUSUBI\n- π **Report issues** - Help us improve\n- π‘ **Suggest features** - We value your ideas\n- π **Share your experience** - Write about MUSUBI on your blog\n\n## License\n\nMIT License - see [LICENSE](LICENSE) for details.\n\n## Credits\n\nMUSUBI synthesizes features from:\n\n- **musuhi** - 20-agent system, steering, EARS format\n- **OpenSpec** - Delta specs, brownfield support\n- **ag2** (AutoGen) - Multi-agent orchestration\n- **ai-dev-tasks** - Simplicity, progressive complexity\n- **cc-sdd** - P-label parallelization, validation gates\n- **spec-kit** - Constitutional governance, test-first\n\n## π Learn More\n\n- [π Documentation](docs/)\n- [π SRS v3.0.0](docs/requirements/srs/srs-musubi-v3.0.0.md)\n- [π
Project Plan v3.0.0](docs/plans/project-plan-v3.0.0.md)\n- [ποΈ Blueprint](Ultimate-SDD-Tool-Blueprint-v3-25-Skills.md)\n- [π Project Plan](PROJECT-PLAN-MUSUBI.md)\n\n---\n\n\n\n**π MUSUBI** - γγγ³ - Bringing specifications, design, and code together.\n\n[](https://github.com/nahisaho/MUSUBI)\n\nMade with β€οΈ for the AI Coding Community\n\n
\n",
"bytes": 53111,
"sha": "f24aef117b66e0268472f304bc09a963d2bb3b04ae254cd96322457abddd7685",
"repo_slug": "nahisaho/musubi",
"fonte": "repo",
"truncated": false,
"api": "https://agentalog.com/api/listings/skl_nahisaho_musubi_claude_skills_constituti_3652c2f3/readme"
}