A Playful Market Simulation Game
Trading Zoo is a whimsical trading simulation game that brings the world of finance to life through anthropomorphized animal teams. Set in a jungle-themed environment, the game features various animal species as sophisticated trading teams, each with their own specialties and trading strategies.
Players can observe and interact with a vibrant marketplace where animal teams trade exotic goods in real-time. The game combines educational aspects of market dynamics with playful, jungle-themed elements.
- Penguins 🐧 - Cool-headed traders specializing in ice futures
- Monkeys 🐒 - Agile merchants expert in fruit trading
- Foxes 🦊 - Cunning market analysts and strategic traders
- Iguanas 🦎 - Patient long-term investors
- Bananas 🍌 - A staple commodity
- Ice 🧊 - Temperature-sensitive luxury good
- Pineapples 🍍 - Exotic fruit commodity
- Fish 🐟 - For the best sushi
- Pebbles 💎 - Beach treasure
- Real-time trading simulation
- Dynamic price fluctuations
- Random market events affecting prices
- Team performance tracking
- Visual data representation through graphs
- Detailed trade history
- Jungle-themed UI with playful animations
Trading Zoo implements a sophisticated multi-agent trading environment powered by AutoGen, where AI agents engage in market activities through a central trading engine. Each animal team is represented by an autonomous LLaMA-based agent capable of processing market information and making independent trading decisions.
-
Trading Engine
- Centralized matching engine for order execution
- Real-time orderbook management
- Price discovery mechanism
- Transaction logging and settlement
-
Market Maker Agent
- Provides baseline market liquidity
- Maintains bid-ask spreads
- Helps prevent extreme price volatility
- Facilitates efficient price discovery
-
Trading Agents (Animal Teams)
- Each agent receives:
- Current portfolio status
- Real-time orderbook data
- Market event notifications
- Historical price data
- Autonomous decision-making based on:
- Team-specific trading strategies
- Market conditions analysis
- Risk management parameters
- Portfolio objectives
- Each agent receives:
[Market Events] → [Trading Engine] ← [Market Maker]
↓ ↑ ↓
[Order Book] → [Trading Agents] ← [Portfolio Data]
- Built using AutoGen's multi-agent conversation framework
- Each agent operates as an independent entity with its own strategy and goals
- Real-time communication through event-driven architecture
- Asynchronous order processing and matching
- State management for portfolio and transaction history
- Market events are broadcast to all agents
- Agents analyze their portfolio and market data
- Trading decisions are converted to orders
- Orders are submitted to the central trading engine
- Market maker provides counterparty when needed
- Matching engine executes valid trades
- Portfolio and market data are updated
- Trading agents receive execution feedback
This multi-agent system creates a dynamic, self-sustaining marketplace where AI-driven animal teams can engage in realistic trading activities while maintaining market stability through the market maker's presence.
While maintaining a fun and accessible atmosphere, the game teaches fundamental concepts of:
- Supply and demand
- Market dynamics
- Price fluctuations
- Trading strategies
- Risk management
- Portfolio tracking
Trading Zoo is designed for:
- Beginners interested in learning about trading
- Students learning economics
- Anyone interested in gamified financial education
- Casual gamers who enjoy management simulations
The game strikes a balance between educational value and entertainment, using the jungle theme and animal characters to make complex market concepts more approachable and engaging.
Backend:
cd backend
uv sync
uv run game.py
Frontend:
cd frontend
npm i
npm run start