Reality doesn't move. The simulation does.
Two chambers, the same real requirement chain. Reality is CHEW's actual current state and never changes here. Simulation is a temporary sandbox — alter one condition and see the exact, honest difference it makes: what moves, and just as often, what doesn't.
0 of 0 real requirements resolved
Current focus:
0 of 0 real requirements resolved
Current focus:
Impact Comparison
Pick two of this example's real unmet requirements and see, in isolation, what resolving each would actually change.
Reality and simulation, computed by the same real rule — on purpose.
Both chambers start from the exact same live fetch of requirementSequence,
basedOnFacts, and capabilityOverview. Reality never changes; the
Simulation chamber's toggles run through the same scenario-engine.js logic the
Unlock Room uses, including the identical "first unresolved requirement, in CHEW's real
order, becomes current focus" rule the real engine documents in
lib/intelligenceEngine.js. Resolving a requirement that isn't already blocking
anything upstream will honestly show no change to current focus — CHEW doesn't
pretend order doesn't matter. Capability-coverage impact is only shown where a requirement
actually links to a real capability in this example; where none do, that's stated plainly
instead of a invented percentage. Nothing here writes to a database or calls CHEW's real
completion path.