Map the result:

Darwin = macOS
Linux = Linux
MINGW* or CYGWIN* or Windows = Windows

Print the detected OS to the user.

Step 2: Create Virtual Environment

Create a Python virtual environment in the current working directory:

macOS / Linux:

python3 -m venv venv
source venv/bin/activate
pip install --upgrade pip

Windows:

python -m venv venv
venv\Scripts\activate
pip install --upgrade pip

If the user specified a Python version argument, use that instead of python3:

$PYTHON_VERSION -m venv venv

Step 3: Install TA-Lib System Dependency

TA-Lib requires a C library installed at the OS level BEFORE pip install ta-lib.

macOS:

brew install ta-lib

Linux (Debian/Ubuntu):

sudo apt-get update
sudo apt-get install -y build-essential wget
wget http://prdownloads.sourceforge.net/ta-lib/ta-lib-0.4.0-src.tar.gz
tar -xzf ta-lib-0.4.0-src.tar.gz
cd ta-lib/
./configure --prefix=/usr
make
sudo make install
cd ..
rm -rf ta-lib ta-lib-0.4.0-src.tar.gz

Linux (RHEL/CentOS/Fedora):

sudo yum groupinstall -y "Development Tools"
wget http://prdownloads.sourceforge.net/ta-lib/ta-lib-0.4.0-src.tar.gz
tar -xzf ta-lib-0.4.0-src.tar.gz
cd ta-lib/
./configure --prefix=/usr
make
sudo make install
cd ..
rm -rf ta-lib ta-lib-0.4.0-src.tar.gz

Windows:

pip install ta-lib

If that fails, download the appropriate .whl file from https://github.com/cgohlke/talib-build/releases and install with:

pip install TA_Lib-0.4.32-cp312-cp312-win_amd64.whl

Step 4: Install Python Packages

Install all required packages (latest versions):

pip install openalgo vectorbt plotly anywidget nbformat ta-lib pandas numpy yfinance python-dotenv tqdm scipy numba nbformat ipywidgets quantstats ccxt duckdb psutil

Step 5: Create Backtesting Folder

Create only the top-level backtesting directory. Strategy subfolders are created on-demand when a backtest script is generated (by the /backtest skill).

mkdir -p backtesting

Do NOT pre-create strategy subfolders.

Step 6: Configure .env File

6a. Check if .env.sample exists at the project root. If it does, use it as a template.

6b. Ask the user which markets they will be backtesting using AskUserQuestion:

Indian Markets (OpenAlgo) — requires OpenAlgo API key
Indian Markets (DuckDB) — direct database loading, no API needed
US Markets (yfinance) — no API key needed
Crypto Markets (CCXT) — optional API key for private data

6c. If the user selected Indian Markets, ask for their OpenAlgo API key:

Ask: "Enter your OpenAlgo API key (from the OpenAlgo dashboard):"
If the user provides a key, store it in .env
If the user skips, write a placeholder

6d. If the user selected Indian Markets (DuckDB), ask for the DuckDB database path:

Ask: "Enter the path to your DuckDB database file (e.g., D:/data/market_data.duckdb):"
Auto-detect format: If the database has a market_data table with symbol, exchange, interval, timestamp columns, it is OpenAlgo Historify format (store as HISTORIFY_DB_PATH). Otherwise store as DUCKDB_PATH.
If the user also has OpenAlgo Historify, ask: "Is this an OpenAlgo Historify database? (y/n)"

6e. If the user selected Crypto Markets, ask if they want to configure exchange API keys:

Ask: "Do you have exchange API keys for authenticated data? (Optional — public OHLCV data works without keys)"
If yes, ask for API key and secret key, store in .env
If no, leave them blank in .env

6f. Write the .env file in the project root directory. Use this template, filling in any keys/paths the user provided:

# Indian Markets (OpenAlgo)
OPENALGO_API_KEY={user_provided_key or "your_openalgo_api_key_here"}
OPENALGO_HOST=http://127.0.0.1:5000

# DuckDB Data Sources (direct database loading - fastest)
# Custom DuckDB (user-created with OHLCV table)
DUCKDB_PATH={user_provided_path or ""}
# OpenAlgo Historify DuckDB (market_data table with epoch timestamps)
HISTORIFY_DB_PATH={user_provided_path or ""}

# Crypto Markets (CCXT) - Optional
CRYPTO_API_KEY={user_provided_key or ""}
CRYPTO_SECRET_KEY={user_provided_key or ""}

6g. Add .env to .gitignore if it exists (never commit secrets):

Scripts use find_dotenv() to automatically walk up and find the single root .env, so no copies are needed in subdirectories.

grep -qxF '.env' .gitignore 2>/dev/null || echo '.env' >> .gitignore

Step 7: Verify Installation

Run a quick verification:

python -c "
import vectorbt as vbt
import openalgo
import plotly
import talib
import duckdb
import anywidget
import nbformat
import quantstats as qs
from dotenv import load_dotenv
print('All packages installed successfully')
print(f'  vectorbt: {vbt.__version__}')
print(f'  plotly: {plotly.__version__}')
print(f'  duckdb: {duckdb.__version__}')
print(f'  nbformat: {nbformat.__version__}')
print(f'  quantstats: {qs.__version__}')
print(f'  TA-Lib: available')
print(f'  python-dotenv: available')
"

If TA-Lib import fails, inform the user that the C library needs to be installed first (see Step 3).

Step 8: Print Summary

Print a summary showing:

Detected OS
Python version used
Virtual environment path
Installed packages and versions
Backtesting folder created (strategy subfolders created on-demand by /backtest)
.env file status (configured with keys / placeholder) — single file at project root
Reminder: "Run cp .env.sample .env and fill in API keys if you skipped configuration"

Important Notes

Never install packages globally — always use the virtual environment
TA-Lib C library installation requires admin/sudo privileges on Linux
On macOS, Homebrew must be installed for brew install ta-lib
If the user already has a virtual environment, ask before creating a new one
The backtesting/ folder is where all generated backtest scripts will be saved
NEVER commit .env files — they contain secrets. Always use .gitignore.
If the user provides an API key during setup, write it directly to .env — do not ask them to edit the file manually
python-dotenv is included in the pip install and must be used by all scripts to load .env

setup

Arguments

Steps

Step 1: Detect Operating System