Standalone structural lab
A scanner, not an oracle.
The core identifies confirmed swing highs and lows, records the first causal break, and tracks the resulting impulse until the opposing structure breaks. The same function runs independently on 1H, 4H, and 1D data; no timeframe votes or looks ahead.
Against 134 BTC waves of 8% or more from 2022–2026, the 1H layer covered 93% of waves. A 30-minute candidate led the 1H confirmation by a median 15 hours, but its 44.8% false-positive rate kept it out of the core.
The negative result matters most: prior bars did not predict BOS direction, and timeframe divergence did not hold up as a predictor. BOS remains useful as a mechanical description of structure, not as an autonomous entry signal.
Path continuity and a small BOS → 1H-confirm → 4H-align effect remain open. An options and gamma overlay is being tested separately, with a decision point fixed for the end of September 2026.