AI Video Short Film Creation Assistant

Problem Addressed

AI short-film workflows often break down at specific points: the story reads well but is not shot-able; screenplay, asset prompts, and lighting details drift apart; Seedance 2.0 segment prompts lack a continuous 15-second timeline, foreground/subject/background layering, and state handoff between segments.

How the Skill Works

The skill outputs a staged production chain: story, screenplay, asset prompts, rendering prompts, and optional cinematic scene prompts.

• Mode A: Story Creation: favors visible actions, light, and sound over internal states, and avoids transitional filler.
• Mode B: Script Adaptation: converts the story into a formatted screenplay with scene, action, dialogue, and effect markers, then checks event coverage and continuity.
• Mode C: Seedance 2.0 Prompts: generates character four-view asset prompts and segmented rendering instructions, typically 15 seconds per segment, with global style, character/scene setup, action timeline, and constraints.
• Mode D: Cinematic Scene Prompts: generates single-frame cinematic prompts focused on compositional pressure, traceable lighting, focal length, and camera position.

The workflow pauses after each stage unless the user explicitly authorizes continuous execution. Every document passes repeated self-checks, capped at five rounds, targeting timeline logic, character consistency, format completeness, AI-like phrasing, and cross-document alignment.

Scope and Caveats

It is best for 1–5 minute sci-fi, atmospheric, suspense-reversal, or fantasy short films, not for long setups, dialogue-heavy scenes, or continuous subtle acting. The output supports AI image/video prompt workflows and does not replace storyboarding, editing, or live-shot blocking; if issues remain after five self-check rounds, the result is flagged for manual review.