Derived by the Research Lead from the operator FX breadth 10 swing-donchian brief: test independent long/short prior-channel breakouts on GBPJPY.FX and EURJPY.FX using close-stamped 4-HOUR bars, targeting 2–15 calendar-day swings. This is a partial-rule completion, not an imported fully specified system. Chosen defaults are a 30-bar entry channel, 12-bar exit channel, fixed 2% protective stop, 15-calendar-day maximum holding time, and up to 50% equity notional per pair. Protective or channel exits may occur before day 2; no minimum holding lock defeats risk control. Orders decided at a completed session bar close execute at the next eligible bar open; protective stops trigger intrabar in the platform, including gap-through fills. No exchange-hosted offline protection is required. Use the FX24_5 calendar and its proven DST-aware bar grid, never forward-fill closures. The code supports Dukascopy 4H history through the reviewed per-pair source-cutover path, but this task has not verified loaded source switches, cutover manifests, actual bar coverage or the deployed cost-context identity. Require those receipts before evaluation; do not silently downgrade to hourly or daily decisions. Historical Dukascopy coverage and paper availability are distinct. Paper admission requires normal evidence/risk/promotion gates and loaded feed acceptance; any later live test is demo-only pending explicit account/risk binding. This output starts no research worker or orders; subsequent jobs retain daily-start/concurrency controls and user-work priority.