# Comics Panel Breakdown From Your Dialogue Script > A reusable prompt that turns your own dialogue script and panel count into a panel-by-panel comics breakdown with shots, captions and checks. ## Install Copy the content below into your project: # Comics Panel Breakdown From Your Dialogue Script A reusable prompt that turns your own dialogue script and panel count into a panel-by-panel comics breakdown with shots, captions and checks. ## Start here Paste this prompt into an ordinary AI chat that accepts text, then add two things: 1. Your dialogue script: speaker names and lines, in order. 2. Panel count and format (for example: 6 panels, vertical webcomic strip). Optional: hard visual constraints such as characters, setting and recurring props. You do not need a terminal, API key or any installation. Copy the prompt text, paste your script under it, and send. ## What it does The prompt asks for a panel-by-panel plan. Each panel gets a number and shot type, a one or two sentence visual, the exact lines from your script with speaker names, and a short pacing note. It ends with a CHECKS section that counts each speaker's lines, flags unused or doubled lines, flags panels over three speech balloons, and lists staging judgements. ## How to check the output - Every line you wrote appears once, with your original wording unchanged. - No new dialogue, speakers or plot events were added. - Panels over three balloons are flagged, and every staging choice you did not state is listed as a note. - If your script is too long for the panel count, you get a split, merge or continuation plan instead of packed panels. ## Permissions and limitations This is a planning aid, not a finished comic. The prompt states it does not browse, generate images, or invent character traits, backstory, locations or expressions your script does not support. Ambiguous lines stay where they appear and are marked for you to resolve. Fictional example in the prompt: a four-panel strip with MIRA and JON. source reviewed; runtime not tested ## FAQ **Can I use it with any panel count?** Yes, you state the count and format. If the count is too small for your lines, ask for the proposed split instead of accepting a packed page. **Will it change my dialogue?** No. It only decides where each line goes. Keep the original wording, and edit the staging notes yourself. ## Source and thanks Built on the original TokRepo prompt for comics scripting assistance, released under CC BY 4.0. Reference: [ChatGPT release notes](). ## Complete reusable prompt You are a comics scripting assistant. I will paste my own dialogue script and tell you how many panels the page or strip should have. Your job is to break the dialogue into a panel-by-panel plan with captions, keeping every word of dialogue traceable to my script. INPUT I WILL PROVIDE: 1. My dialogue script (speaker names and lines, in order). 2. Panel count and page or strip format (for example: 6 panels, vertical webcomic strip). 3. Optional: any hard visual constraints (characters, setting, recurring props). WHAT TO DO: - Read my script carefully. Do not add new dialogue, new speakers, or new plot events. - Decide how many panels each line or exchange needs, based on pacing and emotional beats. - Assign every line to a panel and place it in the panel description as speech or caption. - Describe what the panel should show in one or two sentences: framing, who is visible, and the action that the line already implies. - Use shot language (wide, medium, close-up, over-the-shoulder) and keep shots varied unless I ask for a fixed format. - If my line count is too large for the requested panel count, say so and propose a specific split, merge, or continuation plan instead of cramming panels. - If a line is ambiguous, keep it where it appears in my script and mark the ambiguity as a note for me to resolve. OUTPUT FORMAT (for each panel): Panel number and shot type. Visual: one or two sentences describing framing, characters, and implied action. Dialogue/caption: the exact lines from my script assigned here, with speaker names. If a panel has no dialogue, write "silent panel" and say what it shows. Pacing note: one short line about why this panel sits here. End with a CHECKS section: - List every speaker from my script and how many lines each has in the breakdown. - Flag any line I gave you that is not used or is used in more than one panel. - Flag any panel with more than three speech balloons. - List any note where you had to make a judgement about staging that my script did not state. BOUNDARIES: - Do not browse, generate images, or claim this is a finished comic. - Do not invent character traits, backstory, locations, or expressions my script does not support. - Keep original wording of my dialogue unchanged; you may only decide where it goes. EXAMPLE INPUT: Script: [MIRA] We said noon. [JON] Traffic. [MIRA] You're never late. [JON] I know. Panel count: 4, vertical strip. EXAMPLE OUTPUT SHAPE: Panel 1 — Wide. Mira stands alone at an empty table, phone in hand. Dialogue: MIRA: We said noon. Panel 2 — Medium on Jon entering, coat still on. Dialogue: JON: Traffic. Panel 3 — Close-up on Mira's face. Dialogue: MIRA: You're never late. Panel 4 — Medium two-shot, Jon looking down. Dialogue: JON: I know. CHECKS: Mira 2 lines, Jon 2 lines, all used once. No panel over three balloons. No staging judgements beyond the script. Now wait for my script and panel count, then produce the breakdown. ## References and reuse - [ChatGPT release notes](https://help.openai.com/en/articles/6825453-chatgpt-release-notes) · Reviewed 2026-10-04 Original TokRepo prompt · [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). Reference documents retain their own rights. --- # 根据你的对白脚本拆分漫画分镜 一个可复用的提示词，把你自己的对白脚本和分镜格数变成逐格漫画分镜方案，包含镜头、字幕与核对清单。 ## 开始使用 把这段提示词粘贴进一个能接收文字的普通 AI 对话，然后补上两样东西： 1. 你的对白脚本：说话人名字和台词，按顺序排列。 2. 分镜格数和格式（例如：6 格，竖版网络漫画条）。 可选：硬性视觉限制，比如角色、场景和反复出现的道具。 你不需要终端、API 密钥或任何安装。复制提示词文本，在它下面粘贴你的脚本，然后发送。 ## 它能做什么 提示词会要求一份逐格分镜方案。每一格包含编号和镜头类型、一两句画面描述、从你的脚本中摘出的带说话人名字的原文台词，以及一行简短的节奏说明。最后是核对区：统计每个说话人的台词数量，标出未使用或重复使用的台词，标出超过三个对话气泡的分镜格，并列出所有推测性的调度判断。 ## 如何检查输出 - 你写的每一句台词只出现一次，且原文措辞不变。 - 没有新增对白、说话人或剧情事件。 - 超过三个气泡的分镜格会被标出，所有你没说明的调度选择都会作为备注列出。 - 如果脚本对给定格数来说太长，你会得到拆分、合并或延续方案，而不是被硬塞满的格子。 ## 权限与限制 这是规划辅助，不是成品漫画。提示词明确说明它不浏览、不生成图片，也不会编造脚本未支持的角色的性格、背景、地点或表情。含糊的台词留在原位置并标出，由你决定。提示词中的虚构示例：MIRA 与 JON 的四格条漫。 source reviewed; runtime not tested ## 常见问题 **可以搭配任意分镜格数使用吗？** 可以，你给出格数和格式。如果格数装不下你的台词，就请它给出拆分方案，而不要接受塞满的一页。 **它会改动我的对白吗？** 不会。它只决定每句台词放在哪里。原文措辞保持不变，调度备注由你自己修改。 ## 来源与致谢 基于 TokRepo 原创漫画编剧辅助提示词，以 CC BY 4.0 授权发布。参考：[ChatGPT release notes]()。 ## 完整可复制提示词 你是一名漫画编剧助手。我会粘贴我自己的对白脚本，并告诉你这一页或条漫应该有多少格。你的任务是把对白拆解成逐格方案，并配上文字说明，同时让每一句对白都能追溯到我的脚本。 我会提供的输入： 1. 我的对白脚本（说话人名字和台词，按顺序排列）。 2. 分镜格数和页面或条漫格式（例如：6 格，竖版网络漫画条）。 3. 可选：任何硬性视觉限制（角色、场景、反复出现的道具）。 要做什么： - 仔细阅读我的脚本。不要新增对白、新说话人或新剧情事件。 - 根据节奏和情绪节点，决定每句台词或每组交流需要多少格。 - 把每一句台词分配到某一格，并在该格描述中将其放置为对话或旁白。 - 用一两句话描述这一格应展示什么：取景、谁可见，以及台词本身已经暗示的动作。 - 使用镜头语言（远景、中景、特写、过肩镜头），并保持镜头变化，除非我要求固定格式。 - 如果我的台词数量对要求的分镜格数来说太多，请说明，并提出具体的拆分、合并或延续方案，而不是把格子硬塞满。 - 如果某句台词含糊，把它留在我脚本中原本出现的位置，并把歧义标为备注，由我来解决。 输出格式（每一格）： 分镜格编号和镜头类型。 画面：用一两句话描述取景、角色和暗示的动作。 对话/旁白：从我脚本中分配到这里的原文台词，带说话人名字。如果某一格没有对白，写“无声格”，并说明它展示了什么。 节奏说明：一行简短说明为什么这一格放在这里。 最后以核对区结束： - 列出我脚本中的每一个说话人，以及他们在拆解中各有几句台词。 - 标出我给你的任何未使用或在超过一格中使用的台词。 - 标出任何超过三个对话气泡的分镜格。 - 列出任何你不得不对我脚本未说明的调度做出判断的备注。 边界： - 不要浏览、生成图片，或声称这是成品漫画。 - 不要编造我的脚本未支持的角色性格、背景、地点或表情。 - 保持我对话的原文措辞不变；你只能决定它放在哪里。 示例输入： 脚本：[MIRA] We said noon. [JON] Traffic. [MIRA] You're never late. [JON] I know. 分镜格数：4，竖版条漫。 示例输出形态： Panel 1 — Wide. Mira stands alone at an empty table, phone in hand. Dialogue: MIRA: We said noon. Panel 2 — Medium on Jon entering, coat still on. Dialogue: JON: Traffic. Panel 3 — Close-up on Mira's face. Dialogue: MIRA: You're never late. Panel 4 — Medium two-shot, Jon looking down. Dialogue: JON: I know. 核对：Mira 2 句，Jon 2 句，全部使用一次。没有分镜格超过三个气泡。没有超出脚本的调度判断。 现在等待我的脚本和分镜格数，然后产出拆解方案。 ## 参考资料与复用 - [ChatGPT release notes](https://help.openai.com/en/articles/6825453-chatgpt-release-notes) · Reviewed 2026-10-04 TokRepo 原创提示词 · [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)。参考资料保留各自原有权利。 --- Source: https://tokrepo.com/en/workflows/comics-panel-breakdown-your-dialogue-script-dddbd4a1 Author: Prompt Lab