mirror of
https://github.com/heygen-com/hyperframes.git
synced 2026-11-14 21:29:34 +00:00
Vance IngallsandClaude Opus 4.6
3154dd66c1
Brief step: 6 questions (audience, emotion, brand assets, light/dark, surface, key takeaway) asked one at a time, each answer informing the next. Skip what the prompt already answers. Verbalized sampling: generate 8-10 mood boards, label each as high/mid/low mode, keep 2+ rare boards. Different creative modes (formal, emotional, unconventional, speculative). Anti-mode-collapse check ensures no two boards share the same template. Unlikely-but-valid board generation. Design picker template: - 3x3 grid layout (9 boards) - 3-frame carousel per card (title, data, CTA scenes) - Procedural canvas backgrounds (particles, voronoi, perlin heatmap, ring-pulse, scanline, digits, sine-mesh, ripple, blobs) - Backgrounds as selectable option in Phase 2 - Sticky canvas in Phase 2 preview - No labels on Phase 1 cards — design speaks for itself - Density matches concept, not a 15-element checklist - Full implementation descriptions in design.md output Prompt expansion: verbalized sampling for choreography — 3 approaches per scene (content-first, emotion-first, unconventional), pick the best. Video composition: added proximity/grouping rule. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>