MoAI Project Language & User Initializer

This skill manages the comprehensive project initialization workflow that was previously handled in the 0-project.md command. It extracts the complex batched question patterns into a reusable, efficient skill that reduces user interactions while maintaining full functionality.

Core Responsibility

Handle all project setup workflows including:

• Language selection (Korean, English, Japanese, Chinese)
• Agent prompt language configuration (English vs Localized)
• User nickname collection (max 20 chars)
• Team mode configuration (GitHub settings, Git workflows)
• Domain selection processes
• Report generation settings with token cost warnings
• MCP server configuration (Figma Access Token setup)

Usage Patterns

First-Time Project Initialization

# Complete setup workflow
Skill("moai-project-language-initializer")

# Executes: Basic Batch → Team Mode Batch (if applicable) → Report Generation → Domain Selection → MCP Configuration (if applicable)

Settings Modification

# Update specific settings
Skill("moai-project-language-initializer", mode="settings")

Team Mode Configuration

# Configure team-specific settings
Skill("moai-project-language-initializer", mode="team_setup")

Language Support Matrix

Language	Code	Conversation Language	Agent Prompt Language	Documentation Language
English	en	English	English (recommended)	English
Korean	ko	한국어	English/Locale choice	한국어
Japanese	ja	日本語	English/Locale choice	日本語
Chinese	zh	中文	English/Locale choice	中文

Team Mode Workflows

Feature Branch + PR Workflow

• Best for: Team collaboration, code reviews, audit trails
• Process: feature/SPEC-{ID} branch → PR review → develop merge
• Settings: spec_git_workflow: "feature_branch"

Direct Commit to Develop Workflow

• Best for: Prototypes, individual projects, rapid iteration
• Process: Direct develop commits (no branches)
• Settings: spec_git_workflow: "develop_direct"

Per-SPEC Decision Workflow

• Best for: Flexible teams, mixed project types
• Process: Ask user for each SPEC
• Settings: spec_git_workflow: "per_spec"

Token Cost Management

Report Generation Costs

Setting	Tokens/Report	Reports/Command	Total Session Tokens	Cost Impact
Enable	50-60	3-5	150-300	Full cost
Minimal	20-30	1-2	20-60	80% reduction
Disable	0	0	0	Zero cost

Agent Prompt Language Costs

Setting	Language	Token Efficiency	Cost Impact
English	English	Baseline	Standard
Localized	Korean/Japanese/Chinese	15-20% more tokens	Higher cost

Configuration Management

The skill automatically manages .moai/config/config.json persistence:

Basic Configuration Structure

{
  "language": {
    "conversation_language": "ko",
    "conversation_language_name": "한국어", 
    "agent_prompt_language": "localized"
  },
  "user": {
    "nickname": "GOOS",
    "selected_at": "2025-11-05T12:00:00Z"
  }
}

Team Mode Additional Configuration

{
  "github": {
    "auto_delete_branches": true,
    "spec_git_workflow": "feature_branch",
    "auto_delete_branches_rationale": "PR 병합 후 원격 브랜치 자동 정리",
    "spec_git_workflow_rationale": "SPEC마다 feature 브랜치 생성으로 팀 리뷰 가능"
  }
}

Report Generation Configuration

{
  "report_generation": {
    "enabled": true,
    "auto_create": false,
    "user_choice": "Minimal",
    "warn_user": true,
    "configured_at": "2025-11-05T12:00:00Z"
  }
}

Domain Selection Configuration

{
  "stack": {
    "selected_domains": ["frontend", "backend"],
    "domain_selection_date": "2025-11-05T12:00:00Z"
  }
}

Error Handling & Validation

Input Validation

• Nickname: Max 20 characters, special characters allowed
• Language selection: Must be from supported languages list
• Domain selection: Multi-select with skip option
• Team settings: Boolean and enum validation

Configuration Validation

• JSON schema validation for config.json
• Backward compatibility checks
• Mode detection validation
• Required field presence checks

Error Recovery

• Graceful degradation for missing config sections
• Default value application for invalid inputs
• Retry mechanisms for failed batch calls
• Rollback capability for partial configurations

Integration Points

With Alfred Commands

• /alfred:0-project: Primary integration point
• /alfred:1-plan: Uses domain selection for expert activation
• /alfred:2-run: Applies language settings to sub-agent prompts
• /alfred:3-sync: Respects report generation settings

With Other Skills

• moai-core-ask-user-questions: Uses TUI survey patterns
• moai-skill-factory: Can be invoked for skill template application
• moai-core-agent-guide: Provides agent lineup based on domains

Configuration Dependencies

• .moai/config/config.json: Primary configuration store
• mode: Determines team vs personal workflow
• github: Team-specific settings
• language: Conversation and prompt language settings

Best Practices

For Users

• Choose English for agent prompts to reduce token costs (15-20% savings)
• Enable Minimal report generation for cost-effective operation
• Configure team settings upfront for consistent workflow
• Select relevant domains for expert agent activation

For Developers

• Use batch patterns to minimize user interactions
• Provide clear token cost warnings before expensive operations
• Validate all inputs before persisting configuration
• Maintain backward compatibility with existing config files

For Team Collaboration

• Use Feature Branch + PR workflow for code review
• Enable auto-delete branches for repository hygiene
• Select appropriate domains for expert agent routing
• Configure consistent language settings across team
• Set up MCP servers with proper authentication (Figma tokens)

MCP Server Configuration

Figma Access Token Setup

When Figma MCP is detected in .claude/settings.json, guide users through token configuration:

Detection Logic

# Check if Figma MCP is configured
settings_path = Path(".claude/settings.json")
if settings_path.exists():
    settings = json.loads(settings_path.read_text())
    figma_configured = "mcpServers" in settings and "figma" in settings["mcpServers"]

Token Setup Workflow

1. Verify Figma MCP Installation: Check for Figma in mcpServers configuration
2. Guide Token Creation: Direct user to Figma developer portal
3. Secure Token Storage: Configure environment variable or .env file
4. Validation: Test Figma MCP connectivity

User Guidance Messages

🔐 Figma Access Token Setup Required

Your project has Figma MCP configured, but needs an access token:

Steps:
1. Visit: https://www.figma.com/developers/api#access-tokens
2. Create a new access token
3. Choose storage method:
   - Environment variable (recommended): export FIGMA_ACCESS_TOKEN=your_token
   - .env file: Add FIGMA_ACCESS_TOKEN=your_token to .env
   - Shell profile: Add to ~/.zshrc or ~/.bashrc

4. Restart Claude Code to activate token

Token Validation

def validate_figma_token():
    """Test Figma MCP connectivity with token"""
    # Try to access Figma files via MCP
    # Return success/failure with guidance

MCP Server Status Checking

Provide users with current MCP server status:

def check_mcp_status():
    """Check all configured MCP servers"""
    servers = {
        "context7": check_context7_mcp(),
        "figma": check_figma_mcp(),
        "playwright": check_playwright_mcp()
    }
    return servers

Implementation Notes

This skill extracts and consolidates the complex initialization logic from the original 0-project.md command (800 lines) into a focused, reusable skill (500 lines) while maintaining:

• Full functionality: All original features preserved
• UX improvements: Batch calling patterns maintained
• Error handling: Comprehensive validation and recovery
• Integration: Seamless compatibility with existing workflows
• Performance: Optimized configuration management

The skill serves as a foundation for project initialization and can be extended with additional configuration patterns as needed.