confluence-engine

A pluggable trading-strategy framework for Python: technical indicators, market-data adapters (crypto/forex/idx), a backtester, and a clean Strategy interface you implement yourself. Bring your own signal logic and scoring — the framework stays out of your way.

Install

pip install confluence-engine

Quickstart

from confluence_engine import (
    AdapterConfig, StrategyConfig, BreakoutDemo,
    get_adapter, calculate_indicators, run_backtest,
)

cfg = StrategyConfig()
df = get_adapter("crypto", AdapterConfig(exchange="binance")).fetch_ohlcv("BTC/USDT", "1h", limit=300)
df = calculate_indicators(df, cfg.ema_fast, cfg.ema_slow, cfg.rsi_length)
print(run_backtest(df, BreakoutDemo(), cfg))

Write your own strategy

from confluence_engine import Strategy, Signal, StrategyConfig
import pandas as pd

class MyStrategy:
    name = "my-strategy"
    def generate_signal(self, df: pd.DataFrame, config: StrategyConfig) -> Signal:
        last = df.iloc[-1]
        if last["rsi"] < 25:
            return {"signal": "BUY", "reason": "deeply oversold", "score": 0.8,
                    "stop_loss": None, "take_profit": None}
        return {"signal": "HOLD", "reason": "wait", "score": 0.0,
                "stop_loss": None, "take_profit": None}

Pass an optional scorer=lambda df, config: 0.0..1.0 to run_strategy / run_backtest; signals scoring below config.confidence_threshold are downgraded to HOLD.

License

MIT