AI Plot to Seedance Video Prompt Assistant

Problem

Novel, script, or short-drama text often lacks shot language, subject consistency, and generation constraints when pasted directly into a video model. This skill turns Chinese or English plot fragments into prompts for Seedance 2.0 All-in-One Reference video mode, rather than generic image prompts.

How It Works

• Element extraction: Uses rules and a lexicon to identify subject, action, scene, time, weather, props, lighting, and mood.
• Structured assembly: Builds a stable prompt in the order technical parameters → subject → action sequence → scene and lighting → camera language → technical constraints → @reference placeholders.
• Video constraints: Emphasizes slow, continuous, stable motion and adds constraints such as facial stability, no flicker, no clipping, and frame-to-frame consistency.
• Reference placeholders: Includes @Image, @Video, and @Audio notes for binding uploaded reference assets.

Boundaries

Best for 1–3 paragraphs suitable for 8–15 second shots. For long or complex plots, input only the core segment. The @Image/@Video/@Audio entries are placeholders and should be replaced with actual Seedance asset references in the UI. Output is Chinese by default; English prompts require an upstream translation step.