SuperAPI AI Directory

pure-white-handdrawn-ppt-infographic

Design & creation

View repository

Updated 10-15 20:12

About this skill

Pure White Hand-Drawn PPT Infographic

Convert the user's new content into a 16:9 horizontal PPT image with a pure white background, hand-drawn blackboard-writing texture, and layered light-color information, and by default call imagegen skill / built-in image_gen tool to generate the image.

Effect Preview

Effect Preview

assets/preview.png This is the 16:9 sample image that the user confirmed as satisfactory when creating this Skill, used to demonstrate the target visual style. When generating new images later, there is no need to copy the preview image's content; only reuse its corresponding visual system and layout language.

Default Settings

  • Aspect Ratio: 16:9 horizontal.
  • Background: Pure white background, must have clean blank space; do not use off-white, gray background, gradients, paper texture, or large background color blocks.
  • Image Generation Method: Use the imagegen skill, which calls the built-in image_gen tool underneath.
  • Default Delivery: When the user asks for an image, image output, illustration, PPT image, or presentation page, organize the final prompt and call image_gen; when the user explicitly asks only for the prompt, do not call image generation.

Applicable Scenarios

  • Explain a methodology, workflow, product concept, or AI Agent / Skills system as a horizontal PPT infographic.
  • Express contrast relationships such as "old way vs. new way," "chaotic stacking vs. clear path," and "open capability vs. fixed process."
  • Explain 3 to 6 steps, modules, constraints, feedback, or a continuous optimization loop.
  • Generate horizontal illustrations for WeChat public accounts, course PPT illustrations, and hand-drawn explanation pages in presentations.

Style Specifications

  • Overall Positioning: Like an original hand-drawn explanation image in a knowledge-sharing PPT, relaxed, clear, and readable, not like a commercial poster or complex UI.
  • Canvas and Blank Space: 16:9 horizontal, large margins, lots of blank space, with the main subject occupying 70%-80% of the middle of the frame; connect content with hand-drawn arrows, dashed lines, and branching paths.
  • Background: Strictly pure white background, preserving a clean sense of breathing room; do not use gray background, beige background, paper texture, shadow base panels, or gradient light effects.
  • Color System: Black and gray are the main colors for lines and text; light blue, light green, light orange, and light yellow are used for labels, rounded cards, paths, or emphasis lines; the colors are transparent and soft, like a light stroke from a marker.
  • Layout Structure: Prefer left-right comparison, center signpost/turning point, horizontal process, branching paths, and central theme with peripheral short labels. The structure should be understandable at a glance, without mixing too many graphic logics.
  • Text Hierarchy: Place the hand-drawn large title at the top or upper left; only place 2 to 6 character short labels in the middle modules; a short conclusion can be placed at the bottom. Avoid long paragraphs and dense small text.
  • Graphic Language: Lightweight icons such as black-and-gray hand-drawn outlines, rounded labels, sticky notes, dashed boxes, arrows, hand-drawn signposts, small light bulbs, small locks, small targets, numbered steps, and branching paths.
  • Texture: Hand-drawn board writing, fine-liner pen, light-color markers, slightly irregular lines; keep it clean and professional, no dirty smudging.
  • Variable content slots: The title, the left and right sets of concepts, 3 to 6 steps or labels, and 1 bottom conclusion can all be replaced; signposts, arrows, light-color labels, and hand-drawn icons serve as the fixed style language.
  • Things to avoid: No logos, watermarks, WeChat official account wording, real people, photos, 3D, glassmorphism, heavy shadows, complex gradients, complete original text, or proprietary copyrighted characters.

Core Workflow

Step 1: Distill the content

Compress the user input into information suitable for expression in a single image:

  • A 4 to 12 character Chinese handwritten title.
  • One core conclusion, placed at the bottom if necessary.
  • 2 contrasting sides, or 3 to 6 process nodes, path choices, constraints, or feedback stages.
  • Keep each node to a 2 to 6 character short label, preferring nouns or verb phrases.

If the user provides a long text, do not move the original text into the image; first rewrite it into structured labels, process nodes, or left-right comparison dimensions.

Step 2: Apply a layout

Choose one structure based on the content:

  • Left-right comparison: On the left, draw the problem, chaos, or old way; in the middle, use a signpost, arrow, or forked path as the transition; on the right, draw the clear process or new way.
  • Horizontal process: 3 to 6 light-color rounded cards arranged from left to right, connected with hand-drawn arrows.
  • Path branching: A central signpost or topic node extends in multiple directions, suitable for discussing choices, strategies, or capability boundaries.
  • Open framework: The large box on the left carries the goal/boundary/path, and the small cards on the right carry rules, feedback, and optimization.
  • Closed-loop iteration: Place 4 to 5 nodes around the central topic to form a cycle of "scenario-boundary-execution-feedback-optimization".

Keep a single image focused on only one core relationship. The image should look like a carefully drawn PPT board page, not a screen full of stickers.

Step 3: Organize the imagegen prompt

Write the prompt in this order:

  1. Specify the image type and ratio: 16:9 horizontal PPT infographic.
  2. Emphasize a pure white background: must be pure white, not off-white, gray background, texture, or gradient.
  3. Describe the hand-drawn style: black-gray lines, light-color markers, rounded labels, dashed arrows, signposts, and lightweight icons.
  4. Describe the theme and structure: title, left-right/process/path relationships, and a short label for each node.
  5. Limit the text: short Chinese labels, clear and readable, no dense small text.
  6. Add negative constraints: no logo, watermark, WeChat official account wording, photos, 3D, complex backgrounds, or proprietary copyrighted characters.

Step 4: Call imagegen

  • When the user explicitly asks to "generate an image / directly output an image / draw one / pair it with an image / PPT image", use the final prompt to call image_gen。
  • If the user only wants the prompt, output only the complete prompt and do not call image generation.

Prompt Template

请生成一张 16:9 横版 PPT 风格图片,风格为:纯白手绘 PPT 信息图。

主题:[用户主题]
标题:[4 到 12 字中文标题]
画面结构:[左右对比 / 横向流程 / 路径分叉 / 开放框架 / 闭环迭代]
主要内容:
- [模块 1:2 到 6 字短标签]
- [模块 2:2 到 6 字短标签]
- [模块 3:2 到 6 字短标签]
- [模块 4:如需要]
- [模块 5:如需要]
底部结论:[可选,一句短结论]

视觉要求:纯白背景,必须干净留白;黑灰手绘描边和手写风中文标题;少量淡蓝、淡绿、淡橙、淡黄半透明马克笔色块;使用圆角标签、便签、虚线箭头、分叉路径、手绘路牌、小灯泡、小锁、小靶心、编号步骤等轻量图标;整体像知识分享 PPT 的原创手绘讲解图,信息清楚、呼吸感强。

文字要求:只使用中文短标签和少量短句,文字清晰可读,不要密集小字,不要遮挡。

限制:不要 Logo;不要水印;不要公众号字样;不要复制参考图具体内容;不要照片;不要 3D;不要玻璃拟态;不要厚重阴影;不要灰色或米色背景;不要纸张纹理;不要复杂渐变;不要专有版权角色。图片比例必须是 16:9。

Image generation rules

  • Generate a single 16:9 horizontal image by default.
  • Before calling, you may briefly provide the final prompt, but do not give a long explanation of the design process.
  • If the user does not provide a specific topic, first infer it from context; if it truly cannot be inferred, then ask one brief question.
  • When generating a series of diagrams, keep the pure white background, hand-drawn black and gray lines, light color labels, icon density, title font size, and whitespace ratio consistent.