MCP Universal Adapter

⚠️ PROJECT STATUS: Work in Progress This project is in early development. Core functionality is being implemented. Contributions welcome!

Transform any API into a fully functional MCP server in seconds, not days.

The Problem

Building MCP (Model Context Protocol) servers manually is slow and repetitive:

73.6k+ stars on awesome-mcp-servers shows massive demand
Each MCP server currently takes 2-5 days to code
There are 15,000+ public APIs, but only ~100 have MCP adapters
Developers write the same boilerplate over and over

The Solution

MCP Universal Adapter automatically generates production-ready MCP servers from any API specification:

# Transform Stripe API into MCP server
mcp-adapt --preset stripe --api-key sk_test_xxx

# Done! Now use in Claude/Cursor:
# "Create a Stripe customer for john@example.com and charge $99"

Time reduction: 5 days → 5 minutes ⚡

🎯 Vision

Make every API in the world accessible to AI by eliminating the manual adapter-writing bottleneck.

One universal tool that unlocks thousands of APIs for the AI ecosystem.

✨ Features (Planned)

Core Capabilities

✅ OpenAPI 3.x Support - Parse any OpenAPI specification
✅ GraphQL Support - Handle GraphQL schemas
✅ REST Discovery - Auto-detect endpoints from base URL
✅ Multiple Languages - Generate Python or TypeScript servers
✅ Authentication - Support for API Key, Bearer, OAuth2

Pre-configured Presets

🔄 Stripe - Complete payments integration
🔄 GitHub - Repository and issue management
🔄 OpenAI - AI model composition
🔄 Slack - Team communication automation
🔄 SendGrid - Email automation

Advanced Features

🔄 HAR Import - Import from browser DevTools
🔄 Postman Collections - Convert Postman to MCP
🔄 Auto-generated Tests - Pytest/Jest test suites
🔄 Live Reload - Hot reload during development
🔄 VS Code Extension - IDE integration

Legend:

✅ In Progress
🔄 Planned

🚀 Quick Start

Note: Installation instructions will be available when v0.1.0 is released.

Installation (Coming Soon)

pip install mcp-universal-adapter

Basic Usage

# Generate from OpenAPI URL
mcp-adapt https://api.example.com/openapi.json \
  --output my-mcp-server

# Use preset for popular APIs
mcp-adapt --preset stripe --api-key sk_test_xxx

# Discover REST endpoints
mcp-adapt https://api.example.com \
  --discover \
  --output example-mcp

Example: Stripe Payments

mcp-adapt --preset stripe --api-key sk_test_YOUR_KEY

Now in Claude/Cursor:

User: "Create a customer for john@example.com and charge $99 for Premium Plan"
AI: [Uses generated MCP server to]
    1. Create Stripe customer
    2. Create payment intent
    3. Process charge
    4. Return confirmation

🏗️ Architecture

API Input (OpenAPI/GraphQL/REST)
         ↓
    Parser Layer
         ↓
  Schema Normalizer
         ↓
   MCP Generator
         ↓
  Code Templates (Jinja2)
         ↓
MCP Server Output (Python/TS)
         ↓
  Validation & Testing

Components

Parsers - Extract API structure from various formats
Normalizer - Unified API representation
Generator - Creates MCP server code
Templates - Language-specific code templates
Validator - Tests generated servers

📚 Documentation

Documentation is being written. Check back soon!

Quickstart Guide - Coming soon
API Reference - Coming soon
Presets Guide - Coming soon
Cookbook - Example recipes

🛣️ Roadmap

Sprint 1 (Days 1-5): Foundation

[ ] Project setup & repository structure
[ ] OpenAPI 3.x parser
[ ] Python MCP generator
[ ] CLI interface (Typer)
[ ] First working demo with Stripe

Sprint 2 (Days 6-10): Polish & Presets

[ ] Error handling & validation
[ ] 5 API presets (Stripe, GitHub, OpenAI, Slack, SendGrid)
[ ] Comprehensive documentation
[ ] v0.1.0 release

Sprint 3 (Days 11-14): Advanced Features

[ ] GraphQL support
[ ] REST endpoint discovery
[ ] HAR file import
[ ] VS Code extension

Future

TypeScript generator
Web playground (Streamlit)
GitHub Action
Docker images
Community presets marketplace

💡 Use Cases

1. Payments (Stripe)

mcp-adapt --preset stripe
# Enable: "Process $99 payment for customer@email.com"

2. GitHub Automation

mcp-adapt --preset github --token ghp_xxx
# Enable: "Create urgent bug issue in myapp repo"

3. Team Communication (Slack)

mcp-adapt --preset slack --workspace myteam
# Enable: "Post 'Deployment done' to #engineering"

4. Custom Internal APIs

mcp-adapt https://api.mycompany.com/swagger.json
# Enable your company's API for AI agents!

🤝 Contributing

We're in early development and would love your help!

Priority Contributions Needed:

API Presets - Add support for popular APIs
TypeScript Generator - Implement TS template
GraphQL Parser - Complete GraphQL support
Documentation - Write guides and examples
Testing - Improve test coverage

How to Contribute:

# 1. Fork the repository
# 2. Clone your fork
git clone https://github.com/YOUR_USERNAME/mcp-universal-adapter.git
cd mcp-universal-adapter

# 3. Create virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# 4. Install dependencies
pip install -e ".[dev]"

# 5. Create feature branch
git checkout -b feature/amazing-feature

# 6. Make changes and test
pytest tests/

# 7. Commit and push
git commit -m "feat: add amazing feature"
git push origin feature/amazing-feature

# 8. Open Pull Request

See CONTRIBUTING.md for detailed guidelines.

🎯 Success Metrics

Launch Goals (First 30 days):

5,000+ GitHub stars
100+ generated MCP servers
20+ API presets
Listed in awesome-mcp-servers

📜 License

MIT License - see LICENSE file for details.

🌟 Star History

📧 Contact

Bartosz Gaca - AI & Automation Strategist

Email: gaca.bartosz@gmail.com
Website: bartoszgaca.pl
GitHub: @gacabartosz

🙏 Acknowledgments

Model Context Protocol - The protocol that makes this possible
Anthropic - For Claude and MCP specification
awesome-mcp-servers - Inspiration and community

Built with ❤️ to democratize AI-API integration

Stop writing adapters. Start building.

MCP Servers