Bull Vision Agent

A FastAPI application that integrates with Telegram using webhooks and OpenAI Agents SDK for AI-powered stock trading assistance, utilizing MCPHub for multiple MCP server management.

Features

- Telegram bot integration with webhook support
- AI-powered stock analysis using OpenAI Agents SDK
- Multiple MCP server integration via MCPHub:
- Stock news analysis
- Volume wall detection
- Real-time stock data analysis
- Market news integration
- Conversation history tracking
- Trading context management
- MongoDB integration for data persistence

Project Structure

├── app/
│   ├── __init__.py
│   ├── main.py           # FastAPI app with MCPHub initialization
│   ├── api/
│   │   ├── __init__.py
│   │   └── telegram_webhook.py  # Telegram webhook endpoint
│   ├── bot/
│   │   ├── __init__.py
│   │   ├── bot.py            # Bot instance and context management
│   │   ├── agent.py          # AI agent implementation with MCP servers
│   │   ├── context.py        # Conversation context and history
│   │   └── telegram_handler.py  # Process incoming messages
│   ├── core/
│   │   ├── __init__.py
│   │   └── settings.py   # Application settings
│   ├── models/
│   │   ├── __init__.py
│   │   └── news.py           # News data models
│   ├── services/
│   │   ├── __init__.py
│   │   └── mongodb_service.py # MongoDB operations
│   └── startup.py        # Startup events

Setup

1. Clone the repository
2. Install dependencies:

   poetry install

   

3. Copy .env.example to .env and fill in the required values:

   cp .env.example .env

   

4. Edit .env with your actual values:
- TELEGRAM_BOT_TOKEN: Your Telegram bot token from @BotFather
- TELEGRAM_WEBHOOK_URL: The public URL where your bot will receive updates
- HOST and PORT: Server configuration
- AZURE_OPENAI_API_KEY: Your Azure OpenAI API key
- AZURE_OPENAI_ENDPOINT: Your Azure OpenAI endpoint
- AZURE_OPENAI_DEPLOYMENT: Your Azure OpenAI deployment name
- AZURE_OPENAI_API_VERSION: Azure OpenAI API version
- MONGO_URI: MongoDB connection string
- MONGO_DB: MongoDB database name

Running the Application

Start the server:

make run

Webhook Setup

1. Make sure your server is publicly accessible
2. The webhook URL should be in the format: https://your-domain.com/api/telegram/webhook
3. The webhook will be automatically registered when the application starts

Available Commands

- /start - Start the bot
- /help - Show help message

Example Queries

You can ask the bot about:
- Stock analysis (e.g., "Analyze AAPL")
- Market news (e.g., "What's the latest news about Tesla?")
- Trading strategies (e.g., "What's your view on the current market?")
- Volume analysis (e.g., "Check volume patterns for MSFT")

Development Commands

make install    # Install dependencies
make run       # Run the application
make test      # Run tests
make lint      # Run linters
make format    # Format the code
make clean     # Clean up generated files
make setup     # Setup development environment
make check     # Run all checks

Documentation

For detailed instructions on:
1. Setting up the AI Chatbot with Telegram and OpenAI Agents SDK, see init_telegram_openai_agent.md
2. Using MCPHub with multiple MCP servers, see create_telegram_chatbot_multi_mcp_server.md