h3-multiverse-portal-ensemble

H3 Multiverse Portal Ensemble

Stage a clean entrance with original characters, a consistent portal, readable blocking, and one final group composition.

Route the Request

• For prompt writing or refinement, return text only.
• Only when the user explicitly asks to generate or render, invoke $h3-video-generation after composing the prompt.
• Use $h3-prompt-writing for H3 structure and read references/style-blueprint.md.

Stage the Arrival

1. Limit the ensemble to three characters. Give each a unique silhouette, color family, gait, and final position.
2. Define one portal design and keep its color, geometry, particle behavior, and location stable.
3. Use sequential blocking: center arrives first, side character second, opposite-side character last.
4. Assign only one characteristic gesture per subject. Avoid simultaneous combat, transformations, and dialogue.
5. Keep the camera at medium-wide or knee-up framing; use a small slow pull-out or push-in, not a frantic orbit.
6. End with all subjects visible, separated, anatomically stable, and looking toward one common off-screen point.
7. Synchronize portal hum, footsteps, sparks, cloth or armor movement, and one musical rise to the final formation.

Mode and Render Rules

• Use T2VA for original characters described in text.
• Multiple character identity references require native Ref2VA. Pass the ordered character images under reference_images and bind each one explicitly with <Picture N> in the prompt.
• Keep the clip to 4–15 seconds; a 10-second trio arrival is the default demonstration.
• Default to 864x480; allow 960x544 if requested. Explicit 768p means 1344x768 with no downgrade.
• Exclude named franchises, copied costumes, logos, watermarks, extra arrivals, duplicate faces, and uncontrolled weapons.

Output

Return the mode, character blocking map, duration/resolution, complete prompt, and any reference limitation. For a real supported render, wait, verify, and return the absolute Portal URL.