Argus

Tracks the Hidden_Divergence strategy — consistently profitable across all runs. Use when analyzing divergence-based entries or risk allocation.

Tracks the ICT_Killzone_Macro strategy — improved to PF 1.300 but not in active portfolio. Use as a reference for killzone-based experiments.

Historical reference for the TrendPullback strategy — removed from active portfolio in V2.0. Retained as a template for what works in trend following.

Drives the Argus iterative learning loop. After every backtest run, invoke argus_learn.py to score, detect regressions, and generate hypotheses. Update skills/references/ when promoting a change.

Tracks the AVWAP_Confluence strategy — now one of the TOP performers in the portfolio. Use when evaluating strategy strengths or considering AVWAP changes.

Historical reference: SR_Breakout_Retest and Donchian_Breakout previously produced zero trades. This has been resolved in V2.0 — both now generate active signals. Retained as a "don't repeat" reference.

Guidelines for portfolio-level risk management in the Argus backtest system. Covers circuit breakers, correlation handling, position sizing, and market regime detection. Use when modifying engine.py, risk_manager.py, or config.yaml.

Guidelines and historical learnings for tuning individual trading strategies. Use when modifying any strategy file in backtest/strategies/ to understand what has been tried, what worked, and what didn't.