SFCC Dev MCP Server

AI-powered development tools for Salesforce B2C Commerce Cloud

🤖 AI Interface Setup

The SFCC Development MCP Server integrates with multiple AI interfaces. Each has specific setup procedures and capabilities.

📁 Get AI instruction files: ai-instructions folder on GitHub

Quick Start: Get immediate access to SFCC documentation and best practices without any credentials by using documentation-only mode:

{
  "mcpServers": {
    "sfcc-dev": {
      "command": "npx", 
      "args": ["sfcc-dev-mcp"]
    }
  }
}

For full functionality including log analysis and system objects, provide SFCC credentials via --dw-json parameter.

GitHub Copilot Setup

Perfect for developers who want SFCC context directly in VS Code with inline suggestions and code completion.

Installation Steps

  1. Add the instruction file: 
     cp ai-instructions/github-copilot/copilot-instructions.md your-sfcc-project/.github/copilot-instructions.md
  2. Configure MCP server (for MCP-compatible tools): 
    {
  "mcpServers": {
    "sfcc-dev": {
      "command": "npx",
      "args": ["sfcc-dev-mcp", "--dw-json", "/path/to/your/dw.json", "--debug", "true"]
    }
  }
}

    Available command-line options:

    • --dw-json <path> - Path to your dw.json configuration file
    • --debug <true|false> - Enable debug logging (optional, defaults to false)

Features Enabled

  • ✅ Inline code suggestions with SFCC context
  • ✅ Auto-completion for SFCC APIs and patterns
  • ✅ Template generation for controllers, hooks, and components
  • ✅ Real-time error detection and fixes
  • ✅ SFCC-aware refactoring suggestions

Best Practices

  • Place the instructions file in every SFCC project
  • Keep the instructions updated with your project-specific patterns
  • Use descriptive commit messages for better context

Troubleshooting GitHub Copilot

  • Configuration not loading: Ensure the file is named exactly copilot-instructions.md and placed in .github/ directory
  • Suggestions not SFCC-aware: Verify the instruction file is in the project root’s .github/ folder
  • MCP integration issues: Check that MCP client configuration matches the server command exactly

Claude Desktop Setup

The most comprehensive setup with full MCP integration for complex debugging and architecture discussions.

Installation Steps

  1. Copy the instruction file to your project: 
    cp ai-instructions/claude-desktop/claude_custom_instructions.md your-sfcc-project/claude-instructions.md
  2. Locate your Claude Desktop config file:
    • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
    • Windows: %APPDATA%\Claude\claude_desktop_config.json
    • Linux: ~/.config/Claude/claude_desktop_config.json
  3. Choose your operating mode:

Documentation-Only Mode (No SFCC credentials needed)

{
  "mcpServers": {
    "sfcc-dev": {
      "command": "npx",
      "args": ["sfcc-dev-mcp"]
    }
  }
}

Full Mode (With SFCC credentials)

{
  "mcpServers": {
    "sfcc-dev": {
      "command": "npx",
      "args": ["sfcc-dev-mcp", "--dw-json", "/path/to/your/dw.json", "--debug", "true"]
    }
  }
}
  1. Create your dw.json file (for Full Mode): 
    {
  "hostname": "your-instance.sandbox.us01.dx.commercecloud.salesforce.com",
  "username": "your-username", 
  "password": "your-password",
  "client-id": "your-client-id",
  "client-secret": "your-client-secret",
  "site-id": "your-site-id"
}

    Note: The site-id field is optional and only needed for site-specific operations.

Features Enabled

  • ✅ Real-time SFCC documentation access
  • ✅ Multi-turn debugging conversations
  • ✅ Architecture review and planning
  • ✅ Complete log analysis (Full Mode)
  • ✅ System object exploration (Full Mode)
  • ✅ Code version management (Full Mode)

Troubleshooting Claude Desktop

  • Server not starting: Check config file syntax with JSON validator
  • Permission errors: Ensure dw.json file permissions are readable
  • Connection issues: Verify SFCC credentials and network access

Cursor Setup

Advanced setup with multiple context-aware rules for modern AI-powered development.

Installation Steps

  1. Copy the rules structure to your SFCC project: 
    cp -r ai-instructions/cursor/.cursor your-sfcc-project/
  2. Configure MCP server (for MCP-compatible extensions): 
    {
  "mcpServers": {
    "sfcc-dev": {
      "command": "npx",
      "args": ["sfcc-dev-mcp", "--dw-json", "/path/to/your/dw.json", "--debug", "true"]
    }
  }
}

Available Cursor Rules

Always Applied

  • sfcc-development.mdc - Core SFCC patterns and conventions

Auto-Applied (Context Aware)

  • sfra-controllers.mdc - Controller development patterns
  • sfra-models.mdc - SFRA model development and architecture patterns
  • hooks-development.mdc - Hook implementation guidelines
  • system-objects.mdc - Data model and attribute patterns
  • testing-patterns.mdc - Testing templates and best practices
  • isml-templates.mdc - ISML template development patterns
  • job-framework.mdc - Job framework development guidelines
  • scapi-endpoints.mdc - SCAPI endpoint development patterns

Manual Application

  • @debugging-workflows - Debugging guidance and log analysis
  • @security-patterns - Security best practices and vulnerability prevention
  • @performance-optimization - Performance tuning and optimization strategies

Features Enabled

  • ✅ Context-aware code completion based on file type
  • ✅ Real-time validation against SFCC APIs
  • ✅ File-aware refactoring across cartridge structures
  • ✅ Security-first development patterns
  • ✅ Automatic rule application based on context

Using Cursor Rules

// Trigger specific rules with @mentions
// @security-patterns - Apply security best practices
// @performance-optimization - Focus on performance
// @debugging-workflows - Get debugging assistance

Troubleshooting Cursor

  • Rules not applying: Ensure .cursor directory is copied to project root
  • Context-aware features not working: Verify file types match rule patterns
  • MCP integration issues: Check MCP client configuration and server connectivity
  • Performance issues: Consider disabling unused rules for better performance

Available Tools Overview

All AI interfaces can access these tools based on configuration mode:

Documentation-Only Mode Tools (15 tools)

  • ✅ SFCC Documentation (5 tools) - Class information, search, and raw documentation
  • ✅ Best Practices Guides (4 tools) - Comprehensive development guides and hook references
  • ✅ Enhanced SFRA Documentation (5 tools) - Complete SFRA ecosystem with 26+ documents
  • ✅ Cartridge Generation (1 tool) - Automated cartridge structure creation

Full Mode Additional Tools (13 tools)

  • ✅ Log Analysis (7 tools) - Real-time error monitoring and pattern search
  • ✅ System Object Definitions (6 tools) - Custom attributes and site preferences

Comparison & Recommendations

Choose Claude Desktop if:

  • 🎯 You need complex multi-turn conversations
  • 🎯 You want full log analysis capabilities
  • 🎯 You prefer comprehensive debugging sessions
  • 🎯 You work on architecture and design decisions

Choose GitHub Copilot if:

  • 🎯 You primarily work in VS Code
  • 🎯 You want inline code suggestions
  • 🎯 You prefer lightweight integration
  • 🎯 You focus on rapid code completion

Choose Cursor if:

  • 🎯 You want modern AI-powered editing
  • 🎯 You need context-aware rule application
  • 🎯 You prefer file-based intelligence
  • 🎯 You want advanced refactoring capabilities

Next Steps