Part of #4930 → Lane A #5 (Bet 6). Fully data-independent and parallelizable — slot anytime.

Scope (incremental)

• Verifiable sim: theater invalidators/stabilizers/timingMarkers are already structured — bind them to feed metrics so branch predictions get resolved and scored like forecasts (the sim earns its own track record; kills the "narrative theater" criticism).
• Persistent actor memory: actor postures/objectives evolve across runs (continuity records exist; today each run mostly resets) — the genuinely valuable MiroFish idea, without the thousand-agent cost.
• Bridge the two engines: a forecast with routeFacilityKey auto-runs the deterministic run-scenario supply-chain worker → "if this lands: top exposed countries/sectors" card with real trade math. Demo-able differentiator, zero new data needed.
• Pro custom scenarios: parameterized user-triggered theater sims via the existing trigger-simulation queue (bounded templates, not free-form) — builds on the 202 async-job contract in feat(forecast): adopt the 202 async-job contract on trigger-simulation (conditional 202/200) #4811.

Done when

At least the verifiable-sim binding lands (sim branches resolved + scored); the other three are independent follow-ups.

Related: #4930, #4811 (trigger-simulation 202 contract), #5007 (scoring machinery).