Trading Algo Generator — AI-Powered Strategy Builder

Configure Generation

Data File (OHLCV CSV)

Upload CSV

Select a data file or upload your own OHLCV CSV

Static Prompt

You are an expert Python developer specializing in algorithmic trading with the backtesting library (https://pypi.org/project/backtesting/).
Your task is to write a complete, robust Python script that performs all steps below, using only the backtesting library for both backtesting and visualization.
1. Data Format & Preparation
Input CSV file: [[csv_file]]

CSV Columns: time,open_price,close_price,high_price,low_price,coin_volume

Use only the last 2 years of data (relative to the latest date in the CSV).

Parse date/time robustly.

Prepare data in a format compatible with the backtesting library.

2. Strategy Implementation
[[strategy_prompt]]

3. Backtest Parameters
Initial balance: $10,000

Position size: Exactly 5% of current portfolio value for each trade (use integer rounding if required).

Only 1 trade open at a time (no pyramiding or overlapping positions).

Transaction fees: 0.03% for opening a position, 0.06% for closing.

Futures trading: Support both LONG and SHORT positions.

Use only the last 2 years of data.

4. Logging & Performance Metrics
Print to the console:

Each trade: open/close time, type (long/short), entry/exit price, size, fees, P&L per trade, running balance after each trade

Total number of trades, win/loss counts, win rate (profit ratio)

Overall P&L at end of backtest

Maximum drawdown, Sharpe Ratio, Profit Factor, and any common performance metrics

5. Visualization
Use ONLY the built-in plotting functionality of the backtesting library (no Plotly, no matplotlib, no Backtrader plotting)

After running the backtest, save the interactive plot as plot.html in the working directory, but do NOT display it to the user (i.e., do not call plot.show() or any function that opens a window; just save the file).

6. Error Handling & Robustness
Implement full error handling for file loading, data parsing, and invalid/missing columns.

The script must be executable end-to-end as a single file with no user modification.

Add clear comments and set all configurable parameters (fees, position size, etc.) at the top of the script.

7. Code Quality
Modular, clean, and well-commented code.

Follow best practices for Python scripting.

Use [[csv_file]] for data file and [[strategy_prompt]] for strategy.

Auto-Generation Instruction

AI will generate a detailed strategy prompt from these instructions.

Console

[ready] Trading Algo Generator ready

[info] Select data file, configure prompts, and click Generate

Status

Attempts

Time (sec)

Lines of Code

</>

Generated Code

▼

Saved: -

Execution Log

▼

Program Output

▼

▲

Interactive Plot

▼

■

Generated Charts

▼

✦

Strategy Prompt

▼

☰

Final Combined Prompt

▼