Backtesting Tool

Crypto strategy backtesting engine — describe strategies in plain English, test across 108 altcoins.

Live

• Frontend: jumptrading.netlify.app/backtesting.html
• Backend API: backtesting-api-6vb0.onrender.com (free tier — may take ~30s to wake up)

Features

• Natural Language Input — Describe strategies in plain English (e.g., "buy when RSI below 30, sell when overbought")
• Multi-Coin Testing — Test across 108 cryptocurrencies simultaneously with comparison table
• Pre-built Strategies — CRT + CISD, RSI oversold, golden cross, MACD crossover, Bollinger bands, and more
• Screenshot Analysis — Upload TradingView chart screenshots, Claude Vision extracts the strategy automatically
• Configurable Parameters — Starting capital, fees, slippage, test period (1W to 5Y)
• Performance Metrics — Total return, Sharpe/Sortino ratio, max drawdown, win rate, profit factor, trade log
• Risk Management — Built-in stop-loss and take-profit parameters

Tech Stack

• Frontend: Single HTML file (ui/index.html) — vanilla JS, dark theme, Canvas charts
• Backend: Python 3.12 HTTP server (web_server.py)
• AI: Claude Vision (screenshot analysis), Claude Haiku (NL strategy translation)
• Data: 108 coins × 5 years of daily OHLCV candles (Binance via CCXT, cached as Parquet)
• Hosting: Netlify (frontend) + Render (backend)

Project Structure

ui/index.html       — Frontend (single HTML file)
web_server.py       — Python backend server (port 8877)
strategies/         — Strategy parser and definitions (incl. CRT+CISD)
vision/             — Claude Vision screenshot analyzer
engine/             — Backtesting engine core (candle-by-candle, no look-ahead bias)
indicators/         — Technical indicator calculations
metrics/            — Performance metric calculations
data/cache/         — Cached OHLCV data (Parquet files)
render.yaml         — Render deployment config

Local Setup

pip install -r requirements.txt
python web_server.py
# Open http://localhost:8877

Set ANTHROPIC_API_KEY in .env for AI features (screenshot analysis, NL translation).

Deployment

• Frontend auto-deploys to Netlify via jumparo1/JumpTools repo
• Backend auto-deploys to Render from this repo (main branch)