CRV and Precog modes

Two session modes, one protocol.

CRV, the assigned target

The target is selected and sealed to your session before you begin. You receive only the target ID: an arbitrary alphanumeric label carrying no information, in the classic coordinate-protocol pattern. You view, seal, get judged, then see the target.

Precog, the future target

No target exists when you view. It is drawn at random only after your session is sealed. The tasking is effectively: describe the image you will be shown at reveal.

This is precognitive targeting, and it is standard practice, not an exotic variant. Targeting works in both directions of time, provided the tasking is anchored to a precise moment, and the feedback moment is a valid anchor. It also has a structural benefit: a target that does not exist yet cannot leak, and there is nothing to remember about a selection that has never been made.

Which to use

Both modes run under the same blinding, sealing, and judging rules. In practice the line between them is thinner than it looks: even present-time sessions routinely pick up near-past and near-future data, which is exactly why every tasking here is anchored in time. Pick one, or alternate. Method is free.