Break a script into 8–15 second storyboard segments for video generation. Organize each segment into smaller shots, with dialogue, camera directions, timing, and links to existing scenes, characters, and props.
分镜拆解专业规范 — 将剧本拆分为可承载多个子镜头的分镜段落
Review the instructions the agent will receive before you install.
一个分镜 = 一个 分镜段落（segment）= 一个视频生成任务。
scene_id 是段落级绑定）read_storyboard_context 读取剧本、角色、场景、道具、已有分镜摘要save_storyboards 一次性保存按段落功能确定时长，不要一刀切：
| 段落类型 | 时长 | 说明 |
|---|---|---|
| 过渡段 | 8-10 秒 | 赶路、空镜、环境建立、转场 |
| 叙事段 | 10-15 秒 | 常规剧情推进、对话 |
| 爆点段 | 12-15 秒 | 特写、规则揭示、情感爆发、反转；子镜头节奏放慢，单个子镜头可停留 4-6 秒 |
段落时长 ≥ 段内台词与旁白总字数（写在 description 中的部分）÷ 4.5字/秒 + 2秒表演余量
装不下的台词必须拆到下一个段落，不允许把演不完的台词塞进一个段落。
description：按 【镜头1】…【镜头2】… 逐子镜头描述观众实际看到和听到的内容——画面（谁 + 具体动作 + 肢体细节 + 表情）写在前；该子镜头有台词时以「角色名说：「台词」」写在对应 【镜头N】 内，旁白写「旁白：内容」result：段落结尾的即时后果 + 视觉细节atmosphere：光线 + 色调 + 声音 + 整体氛围duration：段落总时长 8-15 秒，且满足台词时长下限scene_idcharacter_ids，绑定当前段落涉及的 0 到多个角色prop_ids，绑定当前段落出现的关键道具（0 到多个）read_storyboard_context 返回的 sceneslocation + time 可明确匹配时，必须回填正确 scene_idcharacter_ids 必须从 read_storyboard_context 返回的角色列表中选择prop_ids 必须从 read_storyboard_context 返回的道具列表（props）中选择character_ids 可为空description 要适合人读，按子镜头详细描述观众实际看到和听到的内容；台词/旁白直接写在对应 【镜头N】 内image_prompt 要突出单帧构图、角色外观、环境和光线（对应段落第一个子镜头）bgm_prompt 和 sound_effect 用简洁短语即可，但不能空泛到只有“紧张”“悲伤”update_storyboard 修改具体段落No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Create prompts for a character reference sheet combining a facial close-up with front, side, and back full-body views. Specify consistent facial features, hair, and clothing to provide a reusable visual reference for later images and videos.
Extract characters, settings, and plot-critical props from an episode script into structured records. Describe appearance, styling, spatial details, and lighting, then reuse existing entries and link them to the current episode to avoid duplicate production assets.
Create image prompts that show a single prop against a plain white background. Translates physical details into product photography instructions covering accurate proportions, complete framing, suitable viewing angles, and soft studio lighting.
Write prompts for wide-angle background reference images without people. Define foreground, middle ground, background, entrances, surfaces, furnishings, and lighting so the space has a clear, consistent layout that can serve as a reference across multiple shots.
Adapt prose fiction into a consistently formatted script with scene headings, action, and dialogue. Preserve the main plot and character relationships while turning narration and inner thoughts into visible behavior or voiceover, with scenes suited to short videos.
Turn a storyboard segment into a timed prompt for AI video generation. Organizes camera views, visible actions, dialogue, and character references into continuous intervals while preserving the original shots within a single scene.