The strategic logic (five load-bearing beliefs)

1. The Domain graph is already the right substrate. System, World, Fate, branches, topology, and the canonical record can support more than research and narrative. The opportunity is to make that model inhabitable without creating a second architecture.
2. Context is the gameplay boundary. Multi-camera play becomes real only when a character, location, or artifact receives what it can know—not an omniscient copy of the record. Better context management improves cost, integrity, information asymmetry, and character autonomy at once.
3. MCP is the common controller. Human natural language, manual selection, and agent choice can all terminate in the same typed tool surface. Domain mechanics should extend that surface rather than inventing private action channels.
4. Cadence makes human-agent worlds humane. Turn-based active and resting periods let idle agents and invested humans coexist without rewarding constant presence. The clock is part of the game constitution, not operational plumbing.
5. Specific worlds prove a general engine. The core becomes credible through sharply different playable presets, not through abstract claims of infinite possibility. Mechanics that recur belong in World; mechanics that do not belong in the Domain.

The strategic discipline

Build the smallest compelling World before the marketplace, league, licensed universe, or real-value economy. Preserve the larger horizon in the concept, but make each expansion earn its engine primitive through actual play.