SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsstoryboard-to-seedance-prompt

storyboard-to-seedance-prompt

Design & creation

View repository

Updated 10-15 17:34

About this skill

Text Storyboard to Seedance 2.0 Prompts

Objective

  • Convert the text storyboard of each clip into a more specific and more executable Seedance 2.0 video prompt, and write the output to the same directory as the text storyboard file.

Input and Output

  • Input:文字分镜.txt(containing all clips of this episode) or a one-sentence requirement
  • Output: in the same directory as the input file, 视频_ClipXXX.prompt.txt(one per clip) and the corresponding video Clip.

Workflow

  • View the user-specified 文字分镜.txt, divide by clip and read the storyboard content item by item; do not skip segments or merge them.
  • For each clip, in that 文字分镜.txt 's same directory, create/update 视频_ClipXXX.prompt.txt, first write "Scene and Time", then write the "Storyboard" timing segments, defaulting to fast-paced cutting and covering the full 0–15s duration.
  • Check against the "Hard Standards/Self-Check List" and fill in missing information item by item until the minimum information line and timing completeness are satisfied.
  • Confirm that the last line of each 视频_ClipXXX.prompt.txt is "No BGM, no subtitles".
  • Generate video: use jimeng-skill's multimodal2video to read prompts from 视频_ClipXXX.prompt.txt and generate the clip, and in the generation parameters pass the aspect ratio (16:9/9:16), model (Seedance 2.0), and the character/scene/prop reference image list to maintain consistency.

How to Use

  • For each clip of the text storyboard, fill in according to the scaffolding below, then write the "Execution Instruction Segment", and always append "No BGM, no subtitles" at the end of the file.
  • The prompt text should only contain visuals and execution details; do not write generation parameters such as aspect ratio, model, or reference image list.

Execution Instruction Segment

  • Use 3–6 sentences to cover the full segment's timing; each sentence includes "subject + action + camera language + spatial/lighting details + time point", compressed from the scaffolding.
  • Always append at the end: No BGM, no subtitles.

Hard Standards

  • Do not merge or omit storyboard items; must cover all scenes and actions of the original storyboard.
  • Minimum information line: each clip ≥10 specific nouns, ≥5 verbs, ≥3 camera terms.
  • Do not use vague words such as "beautiful, good-looking, stunning"; instead use measurable details and executable actions.
  • Specify aspect ratio and model: 16:9 or 9:16; Seedance 2.0.
  • Bind consistency assets: character/scene/prop reference image paths and on-screen order, number of appearances.

Self-Check List

  • Are camera position/shot type/camera movement clear
  • Are time points and rhythm complete
  • Are composition and space, lighting and color, and photography parameters filled in
  • Are reference images bound and order/number of appearances annotated
  • Is "No BGM, no subtitles" appended at the end

Generation Tips

  • After writing the 视频_ClipXXX.prompt.txt in the same directory, use the video generation skill to read that file for generation; the reference image list must be passed in to ensure element consistency and visual continuity.

Related Skills

  • text-storyboard (input source)
  • generate-film-video-prompt (writing detail reference)
  • jimeng-skill (video generation, use this skill to generate video)

Example (Video Prompt)

雨夜的老城石板巷道，路灯昏黄，右上角红蓝霓虹断续闪烁，地面积水形成镜面，空气湿冷

分镜:
0–0.8s: 近景开场，黑伞伞沿占画面上半部，雨滴密集成线砸在伞面并沿边缘成串落下；主角下颌与湿润的呼吸白雾在画面下缘闪现，冷蓝霓虹作为主光扫过皮肤纹理。
0.8–2.2s: 中近景平视推近，主角抬眼望向右上方红蓝霓虹，眼神紧绷；路灯在面部右缘形成一条暖黄边光，背景石板路反光呈碎片状闪烁，景深浅、背景拉开虚化。
2.2–3.6s: 低机位近景跟拍脚步，鞋尖踩入水坑溅起 10–15cm 水花，水滴拖影清晰；镜头轻微抖动模拟奔走压迫感，水面映出断续霓虹条状反射被踩碎。
3.6–4.4s: 切到手部特写，伞柄握紧至指节泛白，雨水沿伞柄连续滴落成一条滴线；快门略慢，手部细微颤动出现轻微拖影。
4.4–5.6s: 眼睛极致特写，水滴沿睫毛滑落，瞳孔短促收缩；红蓝霓虹色块在瞳孔镜面中清晰反射，微摇镜捕捉眉梢与眼眶肌肉的紧张抽动。
5.6–7.2s: 中景侧向平移，人物由慢转快，镜头与其同向滑轨侧移；前景路人伞面与肩膀模糊掠过形成自然遮挡，构图稳定在三分线，空间层次拉开。
7.2–8.8s: 低角度大全跟拍建立环境压迫，黑伞半收高举，风衣下摆被横风掀起；镜头短环绕约 30°，巷道两侧霓虹与路灯把墙面切出冷暖交错块面。
8.8–10.2s: 中景反打，巷口拐角在前方，地面霓虹反光形成明显对角线引导；远处警笛红蓝光带掠过一次并在湿墙上滑动，主角略停半拍后再次加速。
10.2–12.6s: 全景拉远，主角沿拐角冲入更暗的巷段并消失在雨幕里；前景旗板与电线形成遮挡制造暗角，雨线更密更斜，地面积水镜面只剩跳动的霓虹反射。
12.6–15s: 全景静止收束，空镜保留雨幕与反射纹理，路灯暖黄光圈在石板路上轻微颤动；最后 1.2s 完全静止，不做额外镜头运动，仅以画面暗示雨声。
不要BGM，不要字幕

Generation Parameters

  • When generating the video, pass in the aspect ratio (16:9/9:16), the model (Seedance 2.0), and the reference image list (character/scene/prop paths) through the video generation skill; these are not written into the prompt text.