← Back

Execution System

RECLAIM

Market Pattern Detection Engine

Paper

RECLAIM is a rule-based trading signal pipeline built to detect a specific market pattern: liquidity sweep → reclaim or market structure shift confirmation → fair value gap → scored setup.

It is not a machine-learning system. In this phase, RECLAIM is a detection and validation engine — not a full trading bot.

The core detection flow uses 5-minute bars as the primary timeframe and 15-minute bars for trend context. The system detects swing highs and lows, identifies a sweep of that level, confirms the sweep with a reclaim or MSS, then looks for a fair value gap. Each result is classified as FORMED or EXPIRED and scored based on trend alignment, session timing, displacement strength, and structure quality.

Every detected setup is written to SQLite along with a full event trail. Validation tooling can rerun replay over any recent window and export CSV reports. The system supports both live mode and replay mode.

Phase 1 covers: deterministic signal detection, scoring, persistence, replay, live orchestration, CLI dispatch, and validation. Execution, brokerage actions, and performance tracking are reserved for later phases.

Visuals and live data coming soon.