README.mdā¢6.95 kB
# ONEDeFi - AI-Powered Multi-Chain DeFi MCP Server
## š Project Information
**Primary Contact**: J Madhan - (https://t.me/MadhanJ)
**Team**: Solo
**Project Title**: ONEDeFi - AI-Powered Multi-Chain DeFi MCP Server
## š” One-Sentence Elevator Pitch
ONEDeFi is an AI-powered Model Context Protocol (MCP) server that enables intelligent DeFi operations across Ethereum, Polygon, and Solana with automated portfolio optimization, risk assessment, and yield farming strategies.
## š Detailed Project Description
ONEDeFi revolutionizes DeFi interaction by combining blockchain technology with advanced AI capabilities. The platform serves as a comprehensive MCP server that allows AI agents to perform sophisticated DeFi operations including:
### šÆ Core Features
- **Multi-Chain Portfolio Management**: Real-time tracking across Ethereum, Polygon, and Solana
- **AI Portfolio Doctor**: Health diagnostics with personalized treatment plans
- **Strategy Sommelier**: Wine-themed AI investment strategies based on risk profiles
- **Smart Chat Assistant**: Intelligent DeFi guidance and market insights
- **Automated Yield Optimization**: AI-driven recommendations for maximizing returns
- **Risk Assessment**: Comprehensive portfolio analysis and risk management
### šļø Technical Architecture
- **Backend**: Python Flask with SQLite database
- **Blockchain Integration**: Web3.py for Ethereum/Polygon, Solana SDK for Solana
- **AI Integration**: Comput3 AI API with LLaMA models
- **Protocol Compliance**: Model Context Protocol (MCP) JSON-RPC 2.0
- **Frontend**: Bootstrap 5 responsive design
- **Deployment**: Gunicorn on Replit infrastructure
### š§ DeFi Protocols Supported
- **DEX Operations**: Uniswap V2/V3, SushiSwap, QuickSwap, Raydium, Orca
- **Lending**: Aave V3, Compound V3, Solend
- **Yield Farming**: Liquidity provision and reward farming
- **Staking**: Liquid staking through Lido
## š ļø Installation Steps
1. **Clone the repository**:
```bash
git clone [your-repo-url]
cd onedefi
```
2. **Install dependencies** (handled automatically by Replit):
```bash
uv sync
```
3. **Set up environment variables** (see Environment Variables section)
4. **Run the application**:
```bash
python main.py
```
The application will start on `http://0.0.0.0:5000` and be accessible via Replit's web interface.
## š Environment Variables
Create a `.env` file or set the following environment variables:
```env
# Required for AI features
OPENAI_API_KEY=your_comput3_api_key_here
# Blockchain RPC URLs (optional - uses public RPCs by default)
ETHEREUM_RPC_URL=https://cloudflare-eth.com
POLYGON_RPC_URL=https://polygon-rpc.com
SOLANA_RPC_URL=https://api.mainnet-beta.solana.com
# Development settings
USE_TESTNET=true
DEBUG=true
# Flask settings
FLASK_ENV=development
SECRET_KEY=your_secret_key_here
```
**Note**: For AI features to work, you'll need a Comput3 API key. Set it in the Replit Secrets tab as `OPENAI_API_KEY`.
## š Usage Example
### 1. Web Interface
Navigate to your Replit URL to access the web interface:
- **Dashboard**: Portfolio overview and analytics
- **AI Features**: Access Portfolio Doctor, Strategy Sommelier, and Chat Assistant
- **API Docs**: Complete API documentation
### 2. MCP Protocol Usage
```python
import requests
# Analyze portfolio
response = requests.post("https://your-repl-url/mcp", json={
"jsonrpc": "2.0",
"method": "defi.portfolio",
"params": {
"wallet_address": "0x742d35Cc6641C88c4f95bbCdDB96a2b0f0f3f6b7f",
"blockchain": "ethereum"
},
"id": 1
})
```
### 3. AI Features Usage
```python
# Portfolio health check
response = requests.post("https://your-repl-url/api/v1/ai/portfolio-checkup", json={
"wallet_address": "0x742d35Cc6641C88c4f95bbCdDB96a2b0f0f3f6b7f",
"blockchain": "ethereum"
})
# Create investment strategy
response = requests.post("https://your-repl-url/api/v1/ai/create-strategy", json={
"goals": "I want steady 8% returns with low risk",
"wallet_address": "0x742d35Cc6641C88c4f95bbCdDB96a2b0f0f3f6b7f"
})
```
## š Known Issues
1. **Icon Warnings**: Feather icons 'wallet' and 'brain' are not valid - these are cosmetic warnings that don't affect functionality
2. **Testnet Mode**: Currently runs in testnet mode for safety - set `USE_TESTNET=false` for mainnet operations
3. **Rate Limits**: Public RPC endpoints have rate limits - consider using premium RPC providers for production
4. **AI Dependencies**: Some AI features require internet connectivity to Comput3 API
## ā
MCP End-to-End Functionality
**Status**: **Yes ā fully functional**
The MCP server is production-ready with:
- ā
All 8 MCP methods implemented and tested
- ā
JSON-RPC 2.0 compliance verified
- ā
Multi-chain blockchain connections active
- ā
AI integration working with Comput3 API
- ā
Web interface fully operational
- ā
Portfolio analytics and risk assessment functional
- ā
Real-time DeFi protocol integration
**Test Results**:
```
š Web Interface: ā
ALL PASS
š¤ MCP Protocol: ā
PASS
š§ AI Agent: ā
PASS
āļø Blockchain Connections: ā
ALL CONNECTED
```
## š Chains Integrated
- ā
**Ethereum** (Mainnet/Testnet)
- ā
**Solana** (Mainnet/Devnet)
- ā
**Polygon** (Mainnet/Testnet)
## š„ļø Primary Compute Provider
**Comput3** - Used for AI analysis, strategy generation, and chat assistance via their LLaMA model endpoints.
## š License
MIT License - Open source and free to use, modify, and distribute.
## šÆ Additional Information
### š Hackathon Features
This project showcases cutting-edge AI integration in DeFi:
- **Portfolio Doctor**: Medical-themed portfolio diagnostics with visual health scores
- **Strategy Sommelier**: Wine-themed investment strategies with personality descriptions
- **Intelligent Risk Assessment**: AI-powered analysis of DeFi positions
- **Multi-Chain Orchestration**: Seamless operations across 3 major blockchains
### š§ Technical Highlights
- **Production-Ready**: Comprehensive error handling, logging, and security measures
- **Scalable Architecture**: Modular design with clear separation of concerns
- **Real-Time Data**: Live blockchain integration with portfolio tracking
- **AI-Native**: Built from ground up with AI integration as core feature
### š Deployment
- **Platform**: Deployed on Replit with automatic scaling and SSL
- **Uptime**: 99.9% availability target with monitoring
- **Performance**: <200ms API response times
- **Security**: Environment variable management and testnet safety mode
### š Future Roadmap
- Additional blockchain support (BSC, Avalanche)
- Advanced trading strategies and automated rebalancing
- Mobile app integration
- Enhanced risk management tools
---
**Contact**: For questions, issues, or contributions, please reach out via Telegram [@MadhanJ](https://t.me/MadhanJ)
*Built with ā¤ļø for the future of AI-powered DeFi*