YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
🦁 Ultra Elite Scalping v3.2
AI-powered forex scalping signal bot — 6 pairs · 12 indicators · Live news sentiment · Claude AI confirmation
✨ What's New in v3.2
| Feature | v3.1 | v3.2 |
|---|---|---|
| Currency pairs | 3 | 6 |
| Indicators | 10 | 12 (+ Bollinger Bands, Momentum) |
| News sentiment | ❌ | ✅ NewsAPI integration |
| AI confirmation | ❌ | ✅ Claude AI (Anthropic) |
| Repository structure | Single file | Modular package |
New Currency Pairs
AUDUSD— Australian Dollar / US DollarUSDCAD— US Dollar / Canadian DollarEURGBP— Euro / British Pound
📁 Repository Structure
ultra_elite_scalping/
├── main.py ← Entry point (run this)
├── requirements.txt
├── .env.example ← Copy to .env and add your API keys
│
├── config/
│ ├── __init__.py
│ └── settings.py ← All configuration (pairs, thresholds, API URLs)
│
├── core/
│ ├── __init__.py
│ ├── bot.py ← Main orchestrator (UltraEliteBot)
│ ├── indicators.py ← 12-indicator engine (IndicatorEngine)
│ └── scoring.py ← Signal scoring with news + AI bonuses (ScoringEngine)
│
├── apis/
│ ├── __init__.py
│ ├── news_api.py ← NewsAPI client + sentiment parser
│ └── ai_api.py ← Anthropic Claude API client
│
├── utils/
│ ├── __init__.py
│ └── display.py ← All terminal colours & formatted output
│
└── logs/
└── bot.log ← Auto-generated runtime log
🚀 Quick Start
1. Clone & Install
git clone https://github.com/swordenkisk/ultra_elite_scalping.git
cd ultra_elite_scalping
pip install -r requirements.txt
2. Configure API Keys
cp .env.example .env
# Edit .env with your keys:
| Key | Where to get it | Required? |
|---|---|---|
ANTHROPIC_API_KEY |
https://console.anthropic.com | Recommended |
NEWS_API_KEY |
https://newsapi.org (free tier) | Recommended |
ALPHA_VANTAGE_KEY |
https://alphavantage.co | Optional (live prices) |
3. Run
python main.py
⚙️ Configuration
All settings live in config/settings.py and can be overridden via .env:
CYCLE_SECONDS = 45 # seconds between sweeps
SIGNAL_THRESHOLD = 65 # minimum score to fire a signal
NEWS_REFRESH_CYCLES = 3 # refresh news every N cycles
Score Weights (total = 120 pts)
| Component | Points |
|---|---|
| Price Action (EMA crossover) | 25 |
| RSI Extremes | 20 |
| Stochastic | 18 |
| MACD | 15 |
| ADX Trend Strength | 12 |
| News Sentiment Bonus | ±10 |
| Claude AI Confirmation | 0–20 |
🤖 AI Analysis Pipeline
For each pair, when the pre-score is within 15 points of the threshold:
- NewsAPIClient fetches the latest forex/macro headlines
- Sentiment scorer checks headlines for bullish/bearish keywords and returns a ±10 bonus
- AIAnalysisClient sends indicator data + top headline to Claude Sonnet, which returns:
- An integer score 0–20 (confidence boost)
- A 2-3 sentence qualitative analysis
- Final score = technical score + news bonus + AI bonus
If no API keys are configured the bot degrades gracefully: mock headlines and mock AI responses are used, so it always runs.
📰 News Sentiment
Headlines are fetched from NewsAPI with keywords:
forex OR currency OR EUR OR USD OR GBP OR JPY OR AUD OR CAD
Pair-specific filtering uses PAIR_KEYWORDS in config/settings.py — e.g., AUDUSD matches articles mentioning "AUD", "Australian dollar", "RBA", etc.
⚠️ Disclaimer
This software is for educational and research purposes only.
It does not constitute financial advice.
Always test in a paper-trading environment before risking real capital.
Past signal performance does not guarantee future results.
👥 Authors
Ultra Elite Dev Team
Version 3.2.0 — February 2026
Built on:
- Anthropic Claude — AI signal confirmation
- NewsAPI — Live market headlines
- colorama — Terminal colours
- python-dotenv — Environment config
📄 License
MIT — see LICENSE for details.