API to MCP Converter

A powerful VS Code extension that automatically analyzes REST APIs in any repository and converts them to Model Context Protocol (MCP) wrappers. This tool makes any REST API accessible through the MCP framework with automatic authentication handling and AI-powered enhancement capabilities.

🚀 Features

🔍 Intelligent API Discovery

Multi-format Support: Scans workspace for OpenAPI/Swagger specifications (JSON/YAML)
Code Analysis: Detects API endpoints in code files across multiple frameworks
Static Code Scanning: Analyzes source code to extract REST API definitions
AI-Enhanced Analysis: Uses Azure OpenAI to enrich API understanding with business context

🛠️ Advanced MCP Server Generation

Complete Server Implementation: Generates production-ready Node.js MCP servers
Multi-Authentication Support: Handles Bearer tokens, API keys, Basic auth, OAuth2
Auto-build & Validation: Automated compilation and testing pipeline
Test Harness Generation: Includes comprehensive testing infrastructure

🌐 Interactive MCP Tester UI

Swagger-like Interface: Browser-based UI for testing MCP servers
Real-time Server Management: Start, stop, and monitor MCP servers
Live Testing: Interactive tool invocation with parameter validation
Visual Server Explorer: Tree view of all available MCP tools and resources

🧠 Azure AI Integration

Intelligent Code Analysis: AI-powered endpoint detection and enhancement
Business Context Extraction: Understands API purpose and use cases
Smart Documentation: Auto-generates comprehensive API documentation
Configurable AI Models: Supports various Azure OpenAI deployments

🎯 Supported Frameworks & Languages

Backend Frameworks

Node.js: Express.js, Fastify, Koa
Python: Flask, FastAPI, Django REST Framework
Java: Spring Boot, JAX-RS
C#/.NET: ASP.NET Core (Minimal APIs & Controllers)
PHP: Laravel, Symfony
Go: Gin, Echo
Rust: Actix-Web, Rocket

Specification Formats

OpenAPI 3.0/3.1 (JSON/YAML)
Swagger 2.0 (JSON/YAML)
Custom API documentation formats

📦 Installation & Setup

1. Extension Installation

# From VS Code Marketplace
# Search: "API to MCP Converter" by deepusingh-fiber

2. Prerequisites Setup

# Install Node.js (required for MCP server generation)
# Download from: https://nodejs.org/

# Install extension dependencies
npm install

3. Azure AI Configuration (Optional but Recommended)

Open VS Code Settings (Ctrl+,)
Search for "Azure AI"
Configure your Azure OpenAI settings:
- Endpoint: Your Azure OpenAI endpoint URL
- API Key: Your Azure OpenAI API key
- Deployment Name: Your model deployment (e.g., gpt-4)
- API Version: API version (default: 2024-12-01-preview)

🚀 Quick Start Guide

Step 1: Analyze Your Workspace

# Command Palette (Ctrl+Shift+P)
> API to MCP: Analyze Workspace for APIs

# Or click the status bar item: "API to MCP Ready"

Step 2: Generate MCP Wrappers

# From Command Palette
> API to MCP: Setup MCP for Cline

# Or right-click any API in the "API Discovery" tree view
> Generate MCP Wrapper

Step 3: Test Your Generated Servers

# Launch the MCP Tester UI
> API to MCP: Open MCP Tester UI

# Or test individual servers
> API to MCP: Test MCP Wrapper

🎮 Available Commands

Core Commands

Command	Description	Shortcut
`API to MCP: Analyze Workspace for APIs`	Scan workspace for API definitions	-
`API to MCP: Setup MCP for Cline`	Complete MCP setup workflow	-
`API to MCP: Generate MCP Servers for Endpoints`	Generate MCP servers for selected endpoints	-
`API to MCP: Create MCP Wrapper from API`	Interactive wrapper creation	-

Testing & Validation

Command	Description
`API to MCP: Test MCP Wrapper`	Run automated test harness
`API to MCP: Open MCP Tester UI`	Launch browser-based testing interface

AI & Analysis

Command	Description
`API to MCP: Get API Specifications from Selected Files`	AI-powered API analysis
`API to MCP: Configure Azure AI Settings`	Setup Azure OpenAI integration

📁 Project Structure

your-project/
├── mcp-wrappers/                 # Generated MCP servers
│   ├── your-api-mcp-server.js   # Main server file
│   ├── package.json             # Dependencies & scripts
│   ├── test-mcp-server.js       # Test harness
│   ├── test-mcp-mock.js         # Mock testing
│   └── README.md                # Server documentation
├── .api-discovery/              # Analysis cache
│   ├── openapi.synthetic.json   # Synthesized OpenAPI spec
│   └── runtime-cache.json       # Runtime discovery cache
└── DISCOVERY_REPORT.md          # Analysis summary

🧪 Testing Your MCP Servers

1. Automated Test Harness

cd mcp-wrappers
npm install
npm run test:server          # Full test suite
npm run test:server:neg      # Negative auth testing
npm run test:mock           # Mock HTTP testing

2. Interactive MCP Tester UI

Launch via command: API to MCP: Open MCP Tester UI
Browser opens at: http://localhost:3001
Features:
- Visual server explorer
- Interactive tool testing
- Real-time response inspection
- Parameter validation

3. Manual Server Testing

# Set authentication (if required)
export API_KEY="your-api-key"
# or
set API_KEY=your-api-key  # Windows

# Run server directly
node your-api-mcp-server.js

🔐 Authentication Configuration

Environment Variables

# API Key Authentication
export API_KEY="your-api-key"

# Bearer Token Authentication  
export BEARER_TOKEN="your-bearer-token"

# Basic Authentication (username:password)
export BASIC_AUTH="username:password"

# OAuth2 Token
export OAUTH_TOKEN="your-oauth-token"

Windows Environment Variables

set API_KEY=your-api-key
set BEARER_TOKEN=your-bearer-token
set BASIC_AUTH=username:password
set OAUTH_TOKEN=your-oauth-token

📊 Development Scripts

Main Project Scripts

npm run compile              # Compile TypeScript
npm run watch               # Watch mode compilation
npm run build               # Full build including UI
npm run dev                 # Development mode
npm run package             # Create VSIX package
npm run lint                # Code linting
npm run test                # Run tests

MCP Tester UI Scripts

# In mcp-tester-ui/ directory
npm run install-all         # Install all dependencies
npm run dev                 # Development server
npm run build               # Production build
npm run backend             # Backend server only
npm run frontend            # Frontend server only

🔧 Advanced Configuration

Azure AI Settings

{
  "azureAI.endpoint": "https://your-resource.openai.azure.com",
  "azureAI.apiKey": "your-api-key",
  "azureAI.deploymentName": "gpt-4",
  "azureAI.apiVersion": "2024-12-01-preview",
  "azureAI.maxTokens": 30000,
  "azureAI.temperature": 0.3,
  "azureAI.debug": false
}

MCP Server Package.json Template

Generated servers include:

{
  "name": "your-api-mcp-server",
  "version": "1.0.0",
  "type": "module",
  "main": "your_api_mcp_server.js",
  "bin": {
    "your-api-mcp-server": "your_api_mcp_server.js"
  },
  "scripts": {
    "start": "node your_api_mcp_server.js",
    "test:server": "node test-mcp-server.js",
    "test:server:neg": "NO_API_KEY=1 node test-mcp-server.js",
    "test:mock": "node test-mcp-mock.js"
  },
  "dependencies": {
    "@modelcontextprotocol/sdk": "^0.5.0",
    "axios": "^1.7.0"
  }
}

🚀 Publishing MCP Servers

1. Prepare for Publishing

cd mcp-wrappers
npm version patch           # Bump version

2. Publish to NPM

npm publish --access public

3. Global Installation

npm install -g your-api-mcp-server
your-api-mcp-server         # Run globally

🔗 MCP Integration

Cline/Claude Desktop Configuration

{
  "mcpServers": {
    "your_api": {
      "command": "your-api-mcp-server",
      "env": {
        "API_KEY": "your-key-here"
      }
    }
  }
}

VS Code Extension Integration

// Spawn MCP server programmatically
const mcpServer = spawn('node', [wrapperPath], {
  env: { ...process.env, API_KEY: 'your-key' }
});

🐛 Troubleshooting

Common Issues

Issue	Cause	Solution
Server exits immediately	Missing API_KEY env var	Set required environment variable
No tools appear	JSON parse failure in host	Check stdout for pure JSON output
401/403 errors	Invalid credentials	Regenerate/rotate API key
Mock test failures	Unexpected HTTP calls	Update nock definitions
Build failures	Missing dependencies	Run `npm install` in project root
UI won't load	Frontend not built	Run `npm run build-tester-ui`

Debug Mode

Enable detailed logging:

{
  "azureAI.debug": true
}

Check output channels:

"API to MCP Debug" - Extension debugging
Terminal - MCP server logs

🛣️ Roadmap & Future Enhancements

Planned Features

[ ] Auto-generate test harness at wrapper generation time
[ ] Schema-driven validation with improved UX
[ ] Multi-auth strategy injection via config files
[ ] Telemetry & analytics (opt-in)
[ ] GraphQL support for modern APIs
[ ] WebSocket endpoint detection
[ ] API versioning handling
[ ] Bulk server generation for multiple APIs

Performance Improvements

[ ] Incremental analysis with change detection
[ ] Parallel processing for large workspaces
[ ] Smart caching strategies
[ ] Memory optimization for large codebases

📚 Documentation

Comprehensive Guides

User Guide - Complete usage instructions
Troubleshooting Guide - Common issues & solutions
MCP Setup & Integration - Cline/MCP configuration
Development Guide - Extension development
API Reference - Technical specifications

Examples & Templates

Stock Price API Example - Complete implementation
Authentication Patterns - Security implementations
Framework-Specific Guides - Language-specific tips

🤝 Contributing

Development Setup

git clone https://github.com/deepusingh_microsoft/apitomcp.git
cd apitomcp
npm install
npm run compile

Testing

# Run extension in development mode
# Press F5 in VS Code to launch Extension Development Host

# Run unit tests
npm test

# Test with sample projects
npm run test:samples

Contribution Guidelines

Fork the repository
Create a feature branch (git checkout -b feature/amazing-feature)
Commit changes (git commit -m 'Add amazing feature')
Push to branch (git push origin feature/amazing-feature)
Open a Pull Request

📄 License

MIT License - see LICENSE file for details.

🙏 Acknowledgments

Model Context Protocol - Claude/Anthropic for the MCP specification
OpenAPI Initiative - For API specification standards
VS Code Team - For the excellent extension API
Azure OpenAI - For AI-powered analysis capabilities

📞 Support & Community

Getting Help

🐛 Issues: GitHub Issues
💬 Discussions: GitHub Discussions
📧 Email: Support available through GitHub

Version Information

Current Version: 0.0.14
VS Code Compatibility: ^1.103.0
Node.js Requirement: 18+
MCP SDK Version: ^0.5.0

⭐ Star this repository if you find it useful!

🔄 Stay Updated - Watch for releases to get the latest features and improvements.

Note: This extension generates MCP servers that require the Model Context Protocol SDK. Dependencies are automatically managed in generated server directories.

API to MCP Converter

Deepu Singh

API to MCP Converter

🚀 Features

🔍 Intelligent API Discovery

🛠️ Advanced MCP Server Generation

🌐 Interactive MCP Tester UI

🧠 Azure AI Integration

🎯 Supported Frameworks & Languages

Backend Frameworks

Specification Formats

📦 Installation & Setup

1. Extension Installation

2. Prerequisites Setup

3. Azure AI Configuration (Optional but Recommended)

🚀 Quick Start Guide

Step 1: Analyze Your Workspace

Step 2: Generate MCP Wrappers

Step 3: Test Your Generated Servers

🎮 Available Commands

Core Commands

Testing & Validation

AI & Analysis

📁 Project Structure

🧪 Testing Your MCP Servers

1. Automated Test Harness

2. Interactive MCP Tester UI

3. Manual Server Testing

🔐 Authentication Configuration

Environment Variables

Windows Environment Variables

📊 Development Scripts

Main Project Scripts

MCP Tester UI Scripts

🔧 Advanced Configuration

Azure AI Settings

MCP Server Package.json Template

🚀 Publishing MCP Servers

1. Prepare for Publishing

2. Publish to NPM

3. Global Installation

🔗 MCP Integration

Cline/Claude Desktop Configuration

VS Code Extension Integration

🐛 Troubleshooting

Common Issues

Debug Mode

🛣️ Roadmap & Future Enhancements

Planned Features

Performance Improvements

📚 Documentation

Comprehensive Guides

Examples & Templates

🤝 Contributing

Development Setup

Testing

Contribution Guidelines

📄 License

🙏 Acknowledgments

📞 Support & Community

Getting Help

Version Information