Class Scene

Stamp additional meta fields on every composed frame. Used for cross-cutting non-modality meta like acousticDb (sound source amplitude, drives propagation reach) and channelId (channel attribution for chat / multi-party DM frames). Merges with whatever the framework otherwise stamps (timestamp, commandId, causingCommandId, modality). Successive calls accumulate.

Stamp the perception-modality attribution on every composed frame's meta.modality. Used by sensory producers (VocalMixin.say → 'hearing', AetherMixin.tell → 'verbal-esp') so reception-side gating at SensorMixin.filterMessage can drop frames whose modality isn't in the recipient's sensorium. System / log / narrative frames omit this call and deliver unconditionally.

Set a shared payload — applied to any audience that doesn't override.

Compose every queued audience frame, stamp commandId / causingCommandId from the ambient ExecutionContext, and dispatch.

Append shared tags. Merged with per-audience auto-tags.

Frame for everyone inside the actor's contents (excluding the actor itself). Requires the actor to be a Container — used when the actor IS the place (haunted location, ambient narrator).

Frame for everyone in the actor's environment except the actor and any explicit target. Requires the actor to be Containable.

Set the topic. Required before .send().

Frame for the actor. Requires the actor to be a Sensor (else throws — composition error, not user-input error).

Frame for an explicit target. Throws if the target isn't a Sensor.