A repeating time loop can make each collectathon run feel purposeful when the reset changes the next route. Start with one short island path, one moon token, and one gate. The key rule is that the world resets at dawn while the token remains collected.
Define what resets at dawn
Track the player's position and temporary objects as run state. At dawn, return the explorer to the hub and restore temporary objects. Keep a separate persistent flag for the moon token. This boundary makes the loop legible and avoids carrying every detail forward.

Let one collectible change the route
Place the token beyond a short challenge and make it unlock one visible gate at the hub. Before collecting it, the gate stays closed. After the next dawn, the player sees the same hub with the gate open. Keep the reward spatial so the consequence is easy to notice.

Replay one complete loop
The local replay begins with tokenCollected false and gateOpen false. The explorer collects the token, then dawn resets position and temporary state. The persistent flag stays true, so the gate opens. Verify all three facts after reset: hub position restored, temporary state cleared, persistent token retained.
| State | Before dawn | After dawn |
|---|---|---|
| --- | --- | --- |
| Player location | Route | Hub |
| Temporary objects | Active | Reset |
| Moon token | Collected | Retained |
| Hub gate | Closed | Open |

Prompt for an AI game maker
Ask for a compact 3D collectathon with a hub, a short token route, and a dawn reset. Specify which state resets and which single collectible persists. Show the gate before and after the loop. Add an explicit replay button so the same route can be tested repeatedly.
Keep movement, camera tuning, and additional collectibles outside this first check. Add another persistent item only after the reset boundary behaves predictably.
Start a 3D collectathon prototype in SEELE
Frequently asked questions
Should the entire inventory persist?
No. Start with one persistent token flag and reset temporary objects. A narrow rule makes it easier to tell whether the loop worked. Add inventory carryover only after this boundary is reliable, and keep that change isolated from movement and route tests.
What should reset when dawn arrives?
Return the player to the hub and restore temporary route state. Keep the moon token flag so the gate can reflect lasting progress. Include the boundary in the replay output for a clear check, and report both the reset and retained values.
How do I know the loop changed exploration?
Check that the same gate is closed before the token and open after the reset. The changed route is the observable result. Repeating the replay should produce the same gate state each time, which confirms that the rule is deterministic.


