sauravdhanuka/vcp-screener ? reverse-engineered prompt

Build me a Python app that screens NSE stocks using the Volume Contraction Pattern strategy and a mean reversion setup for weak markets.

I want it to pull in a large list of NSE stocks, filter out illiquid names, rank candidates by relative strength, detect VCP style bases with contractions and volume dry up, then show a watchlist of stocks that are close to breakout. It should also switch into a simple mean reversion mode when the market is weak, so it can find oversold bounce setups instead of forcing breakouts in bad conditions.

Please include a clean Streamlit dashboard where I can see the current market regime, top ranked stocks, the VCP score, entry price, pivot level, stop loss, and basic trade status. Add backtesting so I can test the strategy over historical data and see return, drawdown, win rate, and number of trades.

If you need to, look up current docs online for anything that helps with data handling or Streamlit best practices.