Read the content already provided by the user -> step by step have the user make multiple-choice selections -> collect reference images/titles/materials -> generate cover prompts that can be used directly with image generation models according to the corresponding style template.
Important: ask only one question at a time, and wait for the user to answer before asking the next one.
If the user has already provided a certain piece of information in the current message, record it directly and skip the corresponding question. In the end, apply only the style template chosen by the user; do not output a generic template.
Except when the user pastes an article, script, topic selection, or uploads an image, all cover decision questions must be made into numbered multiple-choice questions. Do not use open-ended questions to collect style, person, materials, background, font, color, or title preferences.
Question flow (ask one by one, do not ask all at once)
Question 1: Choose a style
List the following 8 styles and have the user choose one:
- Viral Big-Text Overhead Style — Extra-large title at the top, person centered, suitable for strong hook titles
- Giant-Text Split Impact Style — 2-4 core big words split apart to fill the frame, suitable for concept/identity/ability topics
- Beginner-Friendly Explainer Q&A Style — Black-gray tech poster, blurred interface or real indoor background, question mark and explanatory feel, suitable for "what is/how to understand"
- Tutorial Checklist Style — Green checkmark list + step-by-step feel, suitable for installation, deployment, and getting-started tutorials
- Product Review Ranking Style — Big numbers, arrows, dimension labels, suitable for comparisons, rankings, and reviews
- Recommendation and Endorsement Style — Person giving a thumbs-up or recommendation pose, emoji and warm background, suitable for resources/podcasts/tool recommendations
- Sticker Collage Beginner Style — Person + small cards + product icons + stickers, suitable for "getting started/building/combining"
- Black-Background Efficiency Workflow Style — Dark background + logo wall + checklist, suitable for work methods, efficiency flows, and tool combinations
Question 2: Image 1
Ask the user to upload a person reference image (Image 1), and do not provide a "default image" or "generic blogger image" option.
Only ask:
Please upload Image 1, the person reference image. You may upload 1 clear face photo, or 2-3 photos of the same person to maintain character consistency.
If the user explicitly says they do not want to show their face, you must still require them to upload a partial appearance reference image, such as a hand, partial face edge, or side profile outline. Do not generate a default person out of thin air.
After the user uploads the image, only record it as the Image 1 person reference, and do not ask additional questions about the person's appearance.
Question 3: Person's facial expression and action
Give the user the following options:
- Open-mouthed shock — Mouth open, eyes wide, suitable for reviews/contrast/discovery
- Double thumbs-up — Both hands giving a thumbs-up, suitable for recommendations/endorsements/sharing personal favorites
- Pointing at the title — One hand points to large text or a list, suitable for tutorials/cognition
- Hand on cheek in puzzlement — One hand on cheek, serious and questioning expression, suitable for popular science/question hooks
- Fist raised in confidence — One hand raised in a fist or clenched, suitable for beginner/building/action feel
- Both hands open in explanation — Both hands spread beside the face or in front of the chest, suitable for sharing methods
- Let the model decide
Question 4: Extra assets
Ask the user whether they have extra assets such as image 2, image 3, etc., offering only the following options:
- Have 1 asset image — Image 2 is a software screenshot, product image, Logo, or tool interface
- Have 2 asset images — Image 2 is the main asset, image 3 is the auxiliary asset
- Have 3 or more asset images — Suitable for sticker collage, product review, tool wall
- No asset images — The model generates stickers, icons, lists, arrows, and cards based on the theme
When the user selects 1-3, only ask them to upload image 2, image 3, etc. in order, without additional open-ended follow-up questions. If the purpose of the assets must be specified, give multiple-choice options:
- Product/tool subject
- Tutorial result proof
- Background atmosphere asset
- Sticker collage element
Question 5: Background color tone
Give the user the following options:
- Indoor warm-light background — Like a real room, bookshelf, desktop, suitable for recommendations/experience sharing
- Black-gray tech poster background — Black/dark gray grid, digital annotations, tech poster feel, suitable for AI/tools/concept popular science
- Dark tool wall background — Dark gray/black background + software Logo wall, suitable for efficiency workflows/tool combinations
- Blurred software interface background — Let product screenshots or tool interfaces become the atmosphere
- High-saturation contrasting-color background — Yellow/blue/purple/green strong contrast, suitable for viral strong hooks
- Let the model decide
Question 6: Font style
Give the user the following options:
- Variety-show ultra-bold sans — Default first choice, most like a high-click Xiaohongshu cover
- Comic viral title font — More exaggerated, suitable for shock/review/contrast
- Rounded bold sans — Friendlier, suitable for recommendation/podcast sharing
- Tech-style bold sans — More rugged, suitable for AI tools/efficiency workflows
- Bold title + handwritten doodle small text — The main title is very bold, with a small amount of doodle annotation added beside it
- Let the model decide
Question 7: Font color effect
Give the user the following options:
- Soft light yellow #FDFFA7 fill + thick black stroke — Default first choice, strongest click feel
- Pure white fill + thick black stroke — Clearer on dark backgrounds
- #FDFFA7/white mixed layout + thick black stroke — Use #FDFFA7 for keywords, white for auxiliary words
- #FDFFA7 + white highlight + thick black stroke — More like a viral short-video title
- White text on dark background + #FDFFA7 keywords — Suitable for black-background efficiency workflows and review rankings
- Let the model decide
Question 8: Cover title
Based on the content the user has already provided, distill 3 candidate cover titles and let the user choose only by number:
- Candidate Title A
- Candidate Title B
- Candidate Title C
- Regenerate 3 titles
Do not by default ask open-ended questions like "Do you want to modify the title?" If the user proactively gives a title, use the user's title directly.
Title rules:
- The main title should preferably be 4-10 characters
- A subtitle is allowed, but the main title must be the most eye-catching
- Prefer strongly colloquial hooks, such as "Don't choose blindly", "Build from 0", "Full process made public", "Beginners can get started", "Does it actually work"
- Review-type content must highlight numbers, such as "5 products", "7 dimensions"
- Tutorial-type content must highlight low barriers, such as "Zero foundation", "No command line needed", "Complete beginner"
General Visual Guidelines
All styles must follow:
- Xiaohongshu cover composition must be a 3:4 vertical ratio; other ratios or "close to 3:4" are not accepted
- A real person appears on camera, and the person occupies 35%-55% of the frame, unless the style specifies otherwise
- The edges of the person have a light yellow (preferably #FDFFA7), white, or high-contrast bright outline, creating an obvious cutout feel
- The main title is the first visual priority and must be extra large, clear, and short
- The main title should preferably use soft light yellow #FDFFA7 or pure white, and must have a thick black outline
- Whenever yellow is used for title fill, title keywords, or the backing board/color block/banner behind the title, uniformly specify soft light yellow #FDFFA7; do not use fluorescent yellow, highly saturated bright yellow, golden yellow, or orange-yellow
- Use variety-show-style extra-bold black fonts, comic title lettering, or short-video cover large text
- Common elements: green checkboxes, soft light yellow #FDFFA7 arrows, question marks, exclamation marks, emoji, stickers, small cards, software logos
- The background can be an indoor scene, software interface, blurred product image, dark logo wall, black-gray tech poster, or a background extended from reference materials, but it must not compete with the title
- The image should be high-contrast, high-density, and emotionally strong, with a clickable feel that can be understood at a glance
- All important text, faces, and lists must not extend beyond the safe area
Avoid:
- Do not use minimalist whitespace
- Do not use thin fonts
- Do not use a cold magazine style
- Do not use low-contrast light gray text
- Do not use holographic projections, data streams, or particle effects
- Do not let the title block the person's eyes or mouth
- Do not use fluorescent yellow, highly saturated lemon yellow, golden yellow, or orange-yellow for the title, keywords, or title backing board
Core Requirements for Generating Prompts
After receiving all answers, generate the prompt according to the corresponding style template. The prompt must achieve:
- Specific character pose: clearly state where the person is in the image, where the finger is pointing, and what the expression is
- Clear title hierarchy: write the main title, subtitle, and list text clearly and separately
- Fixed yellow color code: whenever yellow is used for the title, keywords, title backing board, title color block, banner, or arrows, explicitly write #FDFFA7
- Clear spatial relationships: what are the foreground/midground/background, and which occludes which
- Clear material references: directly write "reference image 1/image 2/image 3"; do not restate all the details of the reference images
- Obvious style differences: Each style has its own composition, background, text placement, and decoration logic
- Fixed aspect ratio: Every final prompt must explicitly state "3:4 vertical Xiaohongshu cover"
Style 1: Viral Big-Text Overhead Style
A super-large title at the top presses down on the image, with the person centered or in the lower middle. Suitable for strong hook titles, experience reveals, and tutorial entrances.
Visual Features
- Title: Occupies 35%-45% at the top, soft pale yellow #FDFFA7 or white, thick black outline
- Person: Centered in the lower middle, occupying about 45%, with a strong expression
- Background: Indoor or theme-related scene, slightly blurred
- Accents: A small number of arrows, emoji, stickers to reinforce the emotion
Prompt Template
参考图1的人物五官特征，保持人物一致性，真人半身出镜。
[具体表情]，人物居中偏下，占画面约45%，[具体手势和身体姿态]，眼神看向镜头，人物边缘有浅黄色 #FDFFA7 描边，形成清晰抠图感。
画面顶部放置超大中文标题"[封面标题]"，字体为综艺感超粗黑体，柔和浅黄色 #FDFFA7 填充，粗黑描边，轻微白色高光，文字占据画面上方约40%，有强烈压迫感；如果标题背后需要底板或色块，也使用 #FDFFA7，避免荧光黄或金黄。
如果有副标题，将"[副标题]"放在画面底部，白色粗体，黑色描边，字号小于主标题但仍然醒目。
背景是[背景描述]，轻微虚化，降低细节，让人物和标题成为绝对主体。
画面两侧加入3-5个[主题相关贴纸/emoji/箭头/问号]，只做情绪强化，不抢主标题。
3:4竖版小红书封面，高对比，高密度，强点击感，标题清晰可读，人物脸部不被文字遮挡，所有重要元素不超出安全区。
Style 2: Giant Split-Text Impact Style
Split the title into 2-4 giant word blocks, with the person surrounded by the text or sandwiched in the middle. Suitable for core concepts like "product manager soft skills" and "recording work methods".
Visual Features
- Title: Split into multiple word blocks, almost filling the image
- Person: Sandwiched between the word blocks, with exaggerated movements
- Background: Dark street scene, indoor, blurred work scene
- Overall: Like a short-video cover, strong impact, strong rhythm
Prompt Template
参考图1的人物五官特征，保持人物一致性，真人半身出镜。
[具体表情]，人物位于画面中央，占画面约38%，[具体动作：双手指向上方/双手举起/双手打开]，人物边缘有白色或浅黄色 #FDFFA7 描边。
将中文标题"[封面标题]"拆分成[2/3/4]个巨大词块：[词块1]、[词块2]、[词块3]、[词块4]。
词块使用超粗综艺黑体，柔和浅黄色 #FDFFA7 和纯白交替，全部加粗黑描边。词块分别放在画面上方、左侧、右侧和底部，围绕人物排布，部分文字被人物身体轻微遮挡，形成前后层次；如果词块背后有底板或色块，黄色统一使用 #FDFFA7。
背景是[深色或中性背景描述]，轻微虚化，局部有暗角，让 #FDFFA7/白色大字更突出。
加入少量手绘箭头、短横线、感叹号或英文涂鸦小字，数量控制在3个以内，增加街头感和节奏感。
3:4竖版小红书封面，强对比，超大字视觉锤，人物与文字紧密组合但不遮挡眼睛和嘴巴，不超出安全区。
Style 3: Beginner-Friendly Q&A Explainer Style
Black-gray tech poster, blurred software interface, or real indoor background; the person is serious or puzzled, and question marks and plain-language titles highlight "explaining it clearly for beginners".
Visual Features
- Background: Black-gray tech poster, blurred software interface, real indoor background, or background extended from reference material
- Person: Centered slightly low, chin in hand or looking seriously at the camera
- Title: Large text at the top, like a spoken explanation
- Accents: Large question mark, checkmark audience labels
Prompt Template
参考图1的人物五官特征，保持人物一致性，真人半身或大头出镜。
[具体表情：托腮疑惑/认真解释/微微睁大眼睛]，人物位于画面中下方，占画面约42%，[手部动作：单手托腮/手指轻点脸侧/指向上方标题]，人物边缘有白色或浅黄色 #FDFFA7 描边。
画面上方放置超大中文标题"[封面标题]"，字体为超粗黑体，柔和浅黄色 #FDFFA7 填充，粗黑描边，文字分成2-3行，像在用大白话解释概念；标题背后的黄色底板或色块统一使用 #FDFFA7。
人物旁边放置2个巨大的白色问号，黑色描边，轻微投影，问号围绕人物头部但不遮挡脸。
画面左下或右下放置3条绿色勾选标签：[标签1]、[标签2]、[标签3]，每条是绿色勾选框+白色粗体文字+黑色描边。
背景是[黑灰科技海报/模糊软件界面/真实室内/参考素材延展背景]，略微虚化或压暗，保留科技科普氛围，不抢标题和人物。
3:4竖版小红书封面，标题一眼可读，信息清楚，适合小白科普，不超出安全区。
Style 4: Tutorial Checklist Style
A green checkmark checklist clearly tells users what they can learn. Suitable for deployment, installation, getting started, and full-process reveals.
Visual Features
- Title: Large text at the top or bottom, highlighting the tutorial topic
- Checklist: 3 green checkmark items, placed on the side
- Person: Pointing to the checklist, thumbs up, or explaining excitedly
- Background: Real indoor or tool interface
- A colorful bar can be added at the bottom to reinforce the conclusion
Prompt Template
参考图1的人物五官特征，保持人物一致性，真人半身出镜。
[具体表情]，人物位于画面[左侧/右侧/中下方]，占画面约45%，[具体动作：食指指向清单/竖大拇指/手掌朝向标题]，人物边缘有浅黄色 #FDFFA7 描边。
画面顶部放置超大中文标题"[封面标题]"，字体为超粗黑体，柔和浅黄色 #FDFFA7 填充，粗黑描边，文字紧凑有冲击力；标题背后的黄色底板或色块统一使用 #FDFFA7。
画面[左侧/右侧]竖向排列3条绿色勾选清单：
✅ [清单1]
✅ [清单2]
✅ [清单3]
清单文字为白色粗体，黑色描边，每条间距紧凑，像教程卖点。
如果有图2，将图2作为[软件界面/产品截图/教程结果图]放在人物身后或侧边，轻微倾斜，带投影，作为教程证据。
底部可以加入一条[紫色/柔和浅黄色 #FDFFA7/白色]横幅，写"[底部强化短句]"，字体粗大，形成收尾锤。
背景是[背景描述]，略微虚化，标题、人物、清单三者最醒目。
3:4竖版小红书封面，教程感强，信息密度高，清单清晰可读，所有元素不超出安全区。
Style 5: Product Review Ranking Style
Large numbers, arrows, dimension labels, and product names create a review feel. Suitable for "5 products", "7 dimensions", and "side-by-side reviews".
Visual Features
- Large numbers: Occupy the bottom or middle, such as 5 products, 7 dimensions
- Person: Shocked or serious, pointing to the product/ranking
- Elements: Arrows, product cards, dimension labels
- Background: Dark or tool-related background
Prompt Template
参考图1的人物五官特征，保持人物一致性，真人半身出镜。
[具体表情：张嘴震惊/认真测评/眉毛上扬]，人物位于画面[左下/右下/中央]，占画面约40%，[具体动作：手指指向数字/手掌推向产品卡片/侧身看向榜单]，人物边缘有浅黄色 #FDFFA7 描边。
画面顶部放置中文标题"[封面标题前半]"，字体为柔和浅黄色 #FDFFA7 超粗黑体，粗黑描边。
画面底部放置巨大数字和结果短语"[封面标题后半]"，例如"5款产品"、"7个维度"，数字字号最大，柔和浅黄色 #FDFFA7 填充，黑色描边，占据底部约30%。
用一条粗 #FDFFA7 浅黄色箭头从人物手势指向测评结果，箭头旁边放置小字标签"[测评维度/核心结论]"，白色粗体，黑色描边。
如果有图2/图3，将参考图作为产品卡片或软件截图，围绕人物错落摆放，带轻微倾斜和投影；如果没有素材，则生成[主题相关产品卡片/软件Logo/工具图标]。
背景是[深色/工具界面/模糊工作台]，局部暗角，突出 #FDFFA7/白色标题和数字。
3:4竖版小红书封面，测评感强，数字极醒目，人物表情有反差，所有文字清晰可读，不超出安全区。
Style 6: Recommendation and Endorsement Style
The person gives a thumbs up and makes a friendly recommendation, while the title and emoji create a feeling of "sharing a private favorite" and "you definitely have to try this".
Visual Features
- Person: Thumbs-up or recommendation pose, friendly expression
- Title: #FDFFA7/white large text, near the lower or upper part of the person
- Background: Warm indoor, bookshelf, lifestyle scene
- Accents: Hearts, smiley faces, podcast/tool/book stickers
Prompt Template
参考图1的人物五官特征，保持人物一致性，真人半身出镜。
[具体表情：微笑推荐/双手点赞/认真安利]，人物位于画面中央偏[左/右]，占画面约48%，[具体动作]，眼神看向镜头，人物边缘有柔和浅黄色 #FDFFA7 描边。
画面[顶部/底部]放置超大中文标题"[封面标题]"，字体为超粗圆润黑体，柔和浅黄色 #FDFFA7 和白色组合，粗黑描边，亲切但醒目；标题背后的黄色底板或色块统一使用 #FDFFA7。
在人物旁边加入[主题相关emoji]，例如笑脸、爱心、喇叭、耳机、星星，数量控制在3-5个，围绕人物手势分布。
如果有图2，将图2作为推荐对象的小卡片或封面缩略图，放在人物手边，轻微倾斜，带白色边框和投影。
背景是[温暖室内/书架/桌面/生活化场景]，轻微虚化，有真实分享感。
画面可以加入一行竖排小字"[推荐理由短句]"，白色粗体，黑色描边，贴近人物侧边。
3:4竖版小红书封面，种草感强，亲切可信，高点击感，标题和人物都清晰，不超出安全区。
Style Seven: Sticker Collage Beginner Style
A person combined with multiple small cards, reference images, and icons, suitable for "Multi-Agent Beginner", "Building an AI Squad", and "Setting Up a Workflow".
Visual Features
- Person: centered and slightly large, raising a fist or with a confident expression
- Materials: small cards, screenshots, logos, and tool icons surrounding the person
- Title: split into upper and lower sections, highlighting "Beginner/Building/Newbie"
- Background: indoor or light-colored background with sticker layers overlaid
Prompt Template
参考图1的人物五官特征，保持人物一致性，真人半身出镜。
[具体表情：自信得意/兴奋入门/认真推荐]，人物位于画面中央，占画面约45%，[具体动作：单手举拳/手指向上/手掌托起小卡片]，人物边缘有浅黄色 #FDFFA7 描边。
画面顶部放置大标题"[封面标题前半]"，字体为柔和浅黄色 #FDFFA7 超粗黑体，粗黑描边；画面底部放置大标题"[封面标题后半]"，白色或柔和浅黄色 #FDFFA7，粗黑描边，形成上下夹击；标题背后的黄色底板或色块统一使用 #FDFFA7。
围绕人物放置3-5个贴纸式小卡片：[卡片1]、[卡片2]、[卡片3]、[卡片4]，每张卡片有白色圆角边框、轻微旋转和投影，像手机截图或工具卡片。
如果有图2/图3，将参考图作为最重要的小卡片，放在人物左侧或右侧前景；其余卡片生成[主题相关工具图标/流程图标/头像贴纸]。
右下角或左下角加入一个醒目的主题图标，例如[红色麦克风/纸飞机/AI工具Logo/机器人图标]，带投影。
背景是[背景描述]，保持干净但有层次，贴纸和人物形成拼贴感。
3:4竖版小红书封面，入门感强，信息丰富，卡片不遮挡人物脸，标题清晰可读，不超出安全区。
Style Eight: Black Background Efficiency Workflow Style
Dark background, software logo wall, green checklist, and high-saturation title, suitable for efficiency methods, tool combinations, and workflow reveals.
Visual Features
- Background: dark tool logo wall or blurred software interface
- Title: #FDFFA7/white large text, at the top or upper left
- Checklist: green checkmarks, highlighting efficiency gains
- Person: surprised, excited, or explaining, cut out in the foreground
Prompt Template
参考图1的人物五官特征，保持人物一致性，真人半身出镜。
[具体表情：张嘴兴奋/双手打开/惊讶推荐]，人物位于画面中下方，占画面约45%，[具体动作]，人物边缘有白色或浅黄色 #FDFFA7 描边，和深色背景形成强对比。
背景是深色工具墙：[主题相关软件Logo/工具界面/工作流卡片]以网格或错落方式铺在后景，整体轻微虚化和压暗，不抢人物。
画面顶部放置超大中文标题"[封面标题]"，字体为柔和浅黄色 #FDFFA7 超粗黑体，粗黑描边，局部可以用白色字强调关键词；标题背后的黄色底板或色块统一使用 #FDFFA7。
画面右侧或左侧放置3条绿色勾选清单：
✅ [收益1]
✅ [收益2]
✅ [收益3]
清单文字为白色粗体，黑色描边，像短视频封面的效率卖点。
前景加入一个[主题相关emoji/喇叭/键盘/闪电/工具图标]，靠近人物手势，带投影。
3:4竖版小红书封面，深色高对比，高密度，效率感和工具感强，标题一眼可读，人物脸部清晰，不超出安全区。
General Prompt Writing Guidelines
- Write the scene first, then the style: the order of person, title, materials, background, decorations, and composition should be clear.
- The title must appear exactly as is: do not let the model rewrite a title the user has confirmed.
- Checklist text must be short: 2-6 characters per item, suitable for cover reading.
- Number reference images directly: write "Reference Image 1/Image 2/Image 3", do not retell at length.
- Use fewer negative words: describe positively as much as possible what should appear in the image.
- Keep the A Tuntun style: real person on camera, #FDFFA7/white large text, thick black outline, strong emotion, green checkmarks, sticker feel, high click density.