What problem does it solve?
Writing prompts for AI image and video generation models shot-by-shot is error-prone: characters drift between shots, styles are inconsistent, and downstream tools cannot parse free-form prose. This Skill turns a topic into a strict, machine-parseable shot list that downstream image and video generation skills can consume without ambiguity.
Core Features & Use Cases
- Strict shot-list output: Emits an OVERVIEW block plus numbered SHOT blocks, each containing camera direction, image prompt, video prompt, voiceover, and on-screen text.
- Identity and style anchoring: Copies caller-supplied identity anchors and render styles byte-for-byte into every shot so characters and aesthetics stay consistent across the whole video.
- Dialogue-aware video prompts: Embeds Seedance-compatible dialogue and voiceover tags so generated videos include spoken audio and lip-sync, not just subtitles.
- Use Case: A creator asks for a 50-second, 5-shot vertical short drama about a workplace revenge story; the Skill returns five parseable shot blocks with consistent character descriptions, dialogue tags, and 9:16 aspect-ratio tokens ready for image and video generation.
Quick Start
Ask the agent to generate a 5-shot vertical short-video script about your topic, providing a character identity description and a render style to anchor every shot.
