Revive the host app-status / freshness-monitor + watchdog (STILL dead since Wed 7/8 12:01). Third day dark. app-status.json is frozen at 2026-07-08 12:01 and there have been zero [WATCHDOG] RED/GREEN emails since. Today's new atlas-flow-status verifier covers the a2 scan pipeline — it does not replace the 13-app host watchdog on DESKTOP-AKU8I62, which is still down and hiding every other regression. This was #1 yesterday too.
Confirm the Massive deep-history adapter is live and wire it to exec-backtest. massive.py (the deep-data loader) was modified today at 08:20 — possible movement on the 2-week data wall. Verify the key is set on the box and MassiveBarSource actually feeds exec-backtest-j3p-plays, then re-run the J3P filter/edge verdicts off deep history instead of kti's ~14-day choppy window. Still THE J3P gate.
Reconcile J3P_LIVE / live_setups / churn-today against real 7/9–7/10 fills. Two live days now. Per the trust-report OOS caveat, needs a fills-vs-signals reconciliation before James leans on the GO banner in the moment. Also chase down the PBS-pullback scanner FAIL that atlas-flow-status surfaced.