charieswei
/
TradingAgents
mirror of https://github.com/TauricResearch/TradingAgents.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
TradingAgents/CRYPTO_PHASE1_README.md

8.9 KiB
Raw Blame History

Crypto Market Support - Phase 1 Implementation Complete ✅

Overview

Phase 1 of the crypto market migration has been successfully implemented! The TradingAgents framework now supports cryptocurrency market analysis alongside traditional equities.

What's Been Implemented

1. Data Infrastructure ✅

CCXT Integration (Multi-Exchange Support)

  • File: tradingagents/dataflows/ccxt_vendor.py
  • Features:
    • OHLCV data from 100+ exchanges (Binance, Coinbase, Kraken, etc.)
    • Real-time ticker data
    • Order book depth analysis
    • Recent trades history
    • Exchange-level fundamentals (liquidity, volume)
  • No API key required for public market data

Glassnode Integration (On-Chain Analytics)

  • File: tradingagents/dataflows/glassnode_vendor.py
  • Features:
    • Network health metrics (active addresses, transactions)
    • Exchange flow analysis (inflows/outflows)
    • Whale activity tracking
    • Valuation metrics (NVT ratio, MVRV)
    • Supply profitability analysis
  • Requires: Glassnode API key (paid service)

Messari Integration (Crypto Fundamentals)

  • File: tradingagents/dataflows/messari_vendor.py
  • Features:
    • Asset profiles and tokenomics
    • Market metrics (price, volume, market cap)
    • Crypto news aggregation
    • Supply schedule analysis
    • Project fundamentals (consensus, technology)
  • Works without API key (limited data)

2. Configuration System ✅

Crypto-Specific Config

  • File: tradingagents/crypto_config.py
  • Key Settings: 
    "data_vendors": {
    "core_stock_apis": "ccxt",
    "fundamental_data": "messari",
    "onchain_data": "glassnode"
}
  • Risk Parameters (adjusted for crypto volatility):
    • Max position size: 5% (vs 10% for stocks)
    • Max drawdown: 30% (vs 15% for stocks)
    • Risk multiplier: 3.0x (crypto is 3x more volatile)
  • Asset Tiers:
    • BTC: 20% max position
    • ETH: 15% max position
    • Major altcoins: 5% max
    • Small caps: 2% max

Vendor Abstraction Layer

  • File: tradingagents/dataflows/interface.py (updated)
  • Seamlessly routes to appropriate data vendor based on config
  • Supports fallback vendors if primary fails
  • New category: "onchain_data" for crypto-specific metrics

3. Environment Configuration ✅

Updated .env.example

# Crypto Exchange Keys (optional - for authenticated endpoints)
BINANCE_API_KEY=your_key
BINANCE_API_SECRET=your_secret
COINBASE_API_KEY=your_key
COINBASE_API_SECRET=your_secret
KRAKEN_API_KEY=your_key
KRAKEN_API_SECRET=your_secret

# Crypto Data Providers
GLASSNODE_API_KEY=your_key
MESSARI_API_KEY=your_key

4. Testing Suite ✅

Test Script

  • File: test_crypto_data.py
  • Tests:
    • ✅ CCXT OHLCV data fetching
    • ✅ CCXT ticker and order book
    • ✅ Messari fundamentals and news
    • ✅ Glassnode on-chain metrics (if API key set)
  • Run: python test_crypto_data.py

5. Dependencies ✅

Updated requirements.txt

ccxt              # Multi-exchange crypto data
glassnode         # On-chain analytics
python-dotenv     # Environment variable management

Quick Start

Installation

# 1. Install dependencies
pip install -r requirements.txt

# 2. Set up environment (optional - for premium features)
cp .env.example .env
# Edit .env with your API keys

# 3. Test the implementation
python test_crypto_data.py

Basic Usage

Using CCXT for Crypto Market Data

from tradingagents.dataflows.ccxt_vendor import get_crypto_ohlcv, get_crypto_ticker

# Get Bitcoin OHLCV data
btc_data = get_crypto_ohlcv(
    symbol="BTC/USDT",
    timeframe="1d",
    limit=30,
    exchange="binance"
)
print(btc_data)

# Get current ETH ticker
eth_ticker = get_crypto_ticker("ETH/USDT", "binance")
print(eth_ticker)

Using Messari for Fundamentals

from tradingagents.dataflows.messari_vendor import get_crypto_fundamentals_messari

# Get Bitcoin fundamentals
btc_fundamentals = get_crypto_fundamentals_messari("bitcoin")
print(btc_fundamentals)

Using Glassnode for On-Chain Metrics

from tradingagents.dataflows.glassnode_vendor import get_onchain_metrics

# Get BTC on-chain metrics (requires API key)
btc_onchain = get_onchain_metrics("BTC", days=30)
print(btc_onchain)

Using Crypto Config

from tradingagents.crypto_config import get_crypto_config
from tradingagents.dataflows.config import set_config

# Switch to crypto configuration
crypto_config = get_crypto_config()
set_config(crypto_config)

# Now the framework uses crypto data vendors by default

Architecture Changes

Data Flow Comparison

Before (Stocks Only):

Agent → interface.py → Alpha Vantage/yfinance → Stock Data

After (Multi-Asset):

Agent → interface.py → Route based on config:
                      ├─ Alpha Vantage/yfinance → Stock Data
                      ├─ CCXT → Crypto Market Data
                      ├─ Messari → Crypto Fundamentals
                      └─ Glassnode → On-Chain Metrics

New Data Categories

Category Stock Vendor Crypto Vendor
Price Data yfinance CCXT
Fundamentals Alpha Vantage Messari
News Alpha Vantage Messari
On-Chain N/A Glassnode

What Works Without API Keys

✅ CCXT: All public market data (OHLCV, tickers, order books) ✅ Messari: Basic fundamentals, news, market overview ❌ Glassnode: Requires paid API key for on-chain data

API Rate Limits

Service Free Tier Rate Limit
CCXT ✅ Yes Varies by exchange (~1200/min)
Messari ✅ Yes (limited) 20 requests/min
Glassnode ❌ Paid only 60 requests/min

Testing Results

Run the test suite to validate your setup:

python test_crypto_data.py

Expected output:

===============================================================================
  CRYPTO DATA INFRASTRUCTURE TEST SUITE - PHASE 1
===============================================================================

✅ CCXT: PASSED
✅ MESSARI: PASSED
⚠️  GLASSNODE: SKIPPED (API key required)

Results: 2/2 tests passed

🎉 All crypto data tests passed! Phase 1 implementation complete.

Next Steps (Phase 2-5)

Phase 2: Agent Adaptation (3-4 weeks)

  • Create on-chain analyst agent
  • Update fundamentals analyst prompts for tokenomics
  • Enhance technical analyst with crypto indicators
  • Adapt sentiment analyst for crypto social media

Phase 3: Backtesting Validation (3-4 weeks)

  • Build crypto backtesting engine
  • Validate on historical bull/bear cycles
  • Test multiple asset types (BTC, ETH, altcoins)
  • Calibrate risk parameters

Phase 4: Paper Trading (4-8 weeks)

  • Exchange API integration for trading
  • 24/7 monitoring system
  • Validate execution quality
  • Test emergency procedures

Phase 5: Live Deployment (Ongoing)

  • Start with BTC/ETH only
  • Gradual altcoin expansion
  • Continuous monitoring

File Structure

TradingAgents/
├── tradingagents/
│   ├── dataflows/
│   │   ├── ccxt_vendor.py           # NEW: CCXT integration
│   │   ├── glassnode_vendor.py      # NEW: Glassnode integration
│   │   ├── messari_vendor.py        # NEW: Messari integration
│   │   └── interface.py             # UPDATED: Added crypto routing
│   ├── crypto_config.py             # NEW: Crypto configuration
│   └── default_config.py            # UNCHANGED: Stock config
├── test_crypto_data.py              # NEW: Test suite
├── CRYPTO_MIGRATION_PLAN.md         # Migration roadmap
├── CRYPTO_PHASE1_README.md          # This file
├── requirements.txt                 # UPDATED: Added crypto libs
└── .env.example                     # UPDATED: Added crypto keys

Troubleshooting

Import Errors

# Reinstall dependencies
pip install -r requirements.txt --upgrade

CCXT Connection Issues

# Test exchange connectivity
import ccxt
binance = ccxt.binance()
print(binance.fetch_ticker('BTC/USDT'))

Glassnode 401 Errors

  • Check that GLASSNODE_API_KEY is set correctly
  • Verify your subscription is active
  • Note: Free tier not available for Glassnode

Known Limitations

  1. Glassnode requires paid subscription - On-chain analytics are premium features
  2. CCXT rate limits vary - Each exchange has different limits
  3. Messari free tier is limited - Some endpoints require paid API key
  4. No trading execution yet - Phase 1 is data-only (trading in Phase 4)

Support & Documentation

Contributors

Phase 1 implementation completed on October 7, 2025.

Status: ✅ Phase 1 Complete - Ready for Phase 2 (Agent Adaptation)