The game as a systems argument
Sentinel Radius asks how a civilization would be played if intelligence were already infrastructural. It is not simply a science-fiction RTS with AI decoration. Its controls, economy, territorial logic and failure modes are derived from distributed intelligence.
From Stack to gameplay
| Raynor Stack concept | Sentinel Radius mechanic | Player experience |
|---|---|---|
| Attention as finite energy | Routine is automated inside the Radius. | The player spends attention on the frontier, uncertainty and war. |
| AI as coherence carrier | Agents harvest, build, repair, group and recommend responses. | The player sets intention, policy and doctrine instead of repeating worker commands. |
| Ambience as interface | The Radius communicates state through movement, sound, formation and environmental behavior. | The player reads a living field rather than a dashboard alone. |
| Aura as situated presence | Sentinels accumulate state history, absorbed residue and inherited capabilities. | No two developed Sentinels remain identical. |
| Field as inhabitable coherence | Compute, energy, units and buildings share intelligence only where the field can carry it. | Territory is not merely owned; it is cohered. |
Canonical AI-era RTS loop
The Radius as interface innovation
Delegated coherence
Harvesting, repair, ordinary construction, local defense, resource routing and squad composition are handled by the agent network according to player-set priorities.
Contested coordination
Relays are fragile, latency rises, suggestions become uncertain, and units must balance shared intelligence with local autonomy.
Old-school command
Fog of war, scouting, tactical selection, manual routing, escort, deception and improvisation return beyond reliable field coverage.
Organization becomes visible
When the Radius fails, the base does not simply lose hit points: formations fragment, queues stall, verification weakens and autonomous fallbacks activate.
The playable proposition
The prototype therefore tests a precise claim: automation does not necessarily remove strategy. When automation is local, costly, fallible and spatially bounded, it can move strategy away from routine execution and toward intention, frontier judgment and the architecture of control.