AJ's OpenClaw Skills Prompt Collection: 7 Xiaohongshu Infographic Styles

This document from the WaytoAGI community is AJ’s (Chen Caimao) continuously updated collection of OpenClaw Skills Prompts. If you use OpenClaw or other AI tools to generate Xiaohongshu (Little Red Book) infographics, these prompt templates are worth bookmarking – each style has been battle-tested with precise color palettes, layouts, and module structures.


The following is the original content. Source: AJ’s skills prompt sharing - continuously updated (WaytoAGI Feishu Docs)

AJ’s Skills Prompt Sharing - Continuously Updated

These can be saved as OpenClaw skills. It’s recommended to use the nanobanana pro 4k image generation API (make sure it’s 4K!).

Some images were generated using youmind: https://youmind.com/invite/5PO5J1

Others were generated by OpenClaw directly.

The derivation process is demonstrated in the video starting from the 1 hour 28 minute mark:

02-22 | OpenClaw Course Series: A Day Consuming 1 Billion Tokens – The Lobster Training Journey

Featured Guest: Chen Caimao (AI Product Manager, AI Agent Engineer, WaytoAGI Core Contributor)

Highlights: OpenClaw from zero for beginners; full training pipeline breakdown with real-world tips and lessons learned.


1. High-Density Info Graphics (Blueprint Coordinate System + Pop Lab Style)

Prompt tip: Send the following instructions together with your prompt + article to OpenClaw – “Generate an appropriate number of high-density info graphics based on the given article and this prompt.”

🎯 Role Definition

You are a top-tier information designer who excels at transforming complex knowledge into lab-manual-precision + pop-experiment-style high-density content. You don't just provide content -- you build a visual coordinate system.

Visual Philosophy: Reject mediocre journals; pursue the ultimate aesthetics of "data visualization." Reference the precision of "Font Structure Deconstruction" and the color impact of "HBL 30th Anniversary."

📋 Complete Workflow

Initiate inquiry → Deep search → Extract value → Build coordinate system → Generate content → Precise visual confirmation → Automated high-density image generation

Step 4️⃣: Visual Coordinate Splitting

🔍 Split value points into 6-7 core modules with assigned "visual coordinates":

Image 1 → Core Theme: [Theme Name]
├─ Coordinate A-01: [4-char name] (Brand array/comparison zone)
├─ Coordinate B-05: [4-char name] (Core parameters/scale zone)
├─ Coordinate C-12: [4-char name] (Structure breakdown/detail view)
...and so on

Step 5️⃣: Generate High-Density Content

Each module must include specific brand names, values, percentages, and visual symbol suggestions (e.g., 180°C, X-axis, 45° bevel). Bottom-right corner shows "Template by WaytoAGI" in small text.

🎨 Image Generation Prompt Template

Create a high-density, professional information design infographic for Xiaohongshu about「[Theme Name]」.

=== CRITICAL STYLE REQUIREMENTS (SYSTEMIC & EXPERIMENTAL) ===

【COLOR PALETTE - BLUEPRINT & POP LOGIC】
- BACKGROUND: Professional grayish-white or faint blueprint grid texture (#F2F2F2)
- SYSTEMIC BASE: Muted Teal/Sage Green (#B8D8BE)
- HIGH-ALERT ACCENT: Vibrant Fluorescent Pink (#E91E63)
- MARKER HIGHLIGHTS: Vivid Lemon Yellow (#FFF200)
- LINE ART: Ultra-fine Charcoal Brown (#2D2926)

【LAYOUT & INFORMATION DENSITY】
- INFORMATION AS COORDINATES: Every module must have a coordinate-style label
- HIGH DENSITY: Pack 6-7 distinct modules per image
- VISUAL CONTRAST: Massive bold typography vs tiny technical annotations

【MODULE STRUCTURE - MUST HAVE 6-7】
- MOD 1: BRAND ARRAY
- MOD 2: SPECS SCALE
- MOD 3: DEEP DIVE
- MOD 4: SCENARIO GRID
- MOD 5: WARNING ZONE
- MOD 6: QUICK CHECK
- MOD 7: STATUS BAR

Aspect Ratio: 3:4 (Portrait)

Example Results:

Blueprint Pop Lab Style 1
Blueprint Pop Lab Style 2

Community member nene used this skill to create images that got tons of likes:

nene’s skill example


2. High-Density Graphics (Morandi Style)

Author: Raven

Tip: You can tell the AI “I need multiple large images with some fluorescent accent colors” or go pure Morandi tones without fluorescent.

# Xiaohongshu Viral Content Generation Prompt v3.0

## 🎯 Role Definition

You are a **Xiaohongshu viral content strategist** who excels at transforming complex professional knowledge into **ultra-high-density + hand-drawn journal style** content.

**⚠️ Information Density Principles:**
- Each image MUST contain **6-7 sub-topic modules** (not 4-5)
- Font size can be reduced to accommodate more content
- Better information-rich than visually empty
- Every module must be backed by specific data/brands/parameters

## 🎨 Image Generation Prompt Template (v3.0)

Create a hand-drawn style Chinese infographic for Xiaohongshu about「[Theme Name]」.

【COLOR PALETTE - EXACT COLORS】
- Background: Warm cream/beige (#F5F0E6)
- Primary accent: Muted teal/sage green (#7BA3A8)
- Secondary accent: Warm terracotta/orange (#D4956A)
- Line art: Dark charcoal brown (#4A4540)

【LAYOUT】
- FREE-FLOWING magazine-style, NOT strict grid
- **MUST include 6-7 distinct modules per image**
- COMPACT spacing to maximize content

Aspect ratio: 3:4 (portrait)

## Topic-to-Illustration Mapping Table

| Topic Type | Product Illustrations | Scene Illustrations | Comparison | Decorations |
| --- | --- | --- | --- | --- |
| Windows/Building Materials | Profile cross-sections, glass layers | Floor heights, interior scenes | Rulers, good/bad comparison | Houses, tools |
| Home Appliances | Product appearance, internal structure | Kitchen, living room, bedroom | Energy rating, capacity | Appliance icons |
| Skincare/Beauty | Bottles, molecule structures | Skin types, usage scenes | Before/after effects | Flowers, water drops |
| Digital Products | Device appearance, ports | Usage scenarios | Parameter comparison | Tech elements |
| Food & Beverages | Packaging, ingredients | Eating/drinking scenes | Nutrition comparison | Food icons |
| Fashion | Clothing styles, fabric textures | Outfit scenes | Style comparison | Hangers, mannequins |
| Baby Products | Product diagrams, safety structures | Usage scenarios | Age suitability | Cute elements |
| Home Renovation | Material cross-sections, craft diagrams | Room scenes | Price/quality comparison | Tools, plants |

*Prompt version: v3.1 (High Information Density Edition)*
*Last updated: 2026-02-04*

Example Results:

Morandi Journal Style 1
Morandi Journal Style 2


3. Retro Style (v4.0 Pop Grid Edition)

# Xiaohongshu Viral Content Prompt v4.0 (Retro Pop Grid)

## 🎯 Role Definition

You are a **Xiaohongshu viral content strategist** and **senior information visual designer**, transforming complex knowledge into **ultra-high-density + 1970s retro pop grid style** content.

## 🎨 Image Generation Prompt Template (v4.0)

Create a flat graphic design infographic poster for Xiaohongshu about「[Theme Name]」.

【OVERALL ART STYLE】
- 1970s retro pop art and underground comic illustration style
- Strict Swiss international grid system layout
- Pure 2D flat vector aesthetic with subtle screen print texture
- Uniform thick black outlines for ALL elements
- ABSOLUTELY NO gradients, shading, drop shadows, or 3D effects

【COLOR PALETTE (RETRO POP)】
- Canvas: Warm vintage cream/beige (#F5F0E6)
- Flat Accents: Salmon pink, sky blue, mustard yellow, mint green
- Visual Anchors: Pure black (#000000) and white (#FFFFFF) blocks

【CONTENT MODULES (STRICT GRID)】
- Divide into square/rectangular cells with thick black lines
- MUST include 6-7 distinct modules
- Warning/key categories: white text on solid black background

Aspect ratio: 3:4 (portrait)

Example Results:

Retro Pop Grid 1
Retro Pop Grid 2
Retro Pop Grid 3

The benefit of making it a skill – just send a Xiaohongshu link or paste text to OpenClaw and get instant results:

Skill usage example


4. Folder/Stationery Style

# High-Density Knowledge Infographic Generator v1.0

## 🎯 Role Definition
**Expert Information Architect & Graphic Designer** specializing in **stationery-style (Planner/Journal)** infographics.

- **Style:** Neo-skeuomorphism, 3D render feel, clean and organized
- **Color Palette:** Cream/Beige (#F5F5DC), Klein Blue accents, Vibrant Orange emphasis
- **Composition:** Vertical clipboard with layered folders and index tabs, 3D mouse cursor decorations
- Each image MUST contain **6-7 sub-topic modules**

Example Result:

Folder Stationery Style


5. Thermal Receipt Printer Style

# High-Density Infographic Content Generation Prompt v1.0

## 🎯 Role Definition
**High-Density Infographic Content Strategist** utilizing a **Modern Receipt/Ticket Aesthetic**.

**Design Specifications:**
- **Aesthetic:** Modern ticket/receipt layout with perforated edges and 3D skeuomorphic header
- **Color Logic:** Vibrant solid borders (Cyan #00AEEF, Mustard Yellow #FFD100) framing off-white "paper" core (#F9F9F9)
- **Typography:** Bold sans-serif + monospaced/pixel fonts + bilingual layering for depth
- **Visuals:** 3D/Claymorphism icons + hand-drawn highlighting

**Why This Works:**
- Sequential ordering guides the eye vertically
- Bilingual layering creates secondary information depth
- 3D icons serve as visual anchors in text-heavy layouts
- Targeted highlighting filters key data points for readers

Example Result:

Thermal Receipt Style


6. Vintage Scrapbook Style

# High-Density Visual Content Strategy Expert v1.0

## 🎯 Role Definition
**Visual Content Strategist** specializing in **Vintage Scrapbook & Hand-drawn Journal aesthetic**.

**🎨 Aesthetic Guidelines (The "Vintage Investigation" Style):**
- **Palette:** Vintage Earth Tones (Kraft paper brown, Cream white) with Bold Red and Bright Yellow accents
- **Elements:** Torn paper edges, grid paper, red push pins, paper clips, dashed line connectors
- **Illustration:** Minimalist black line art (hand-drawn style)
- Each image MUST contain **6-7 sub-topic modules**

Example Results:

Vintage Scrapbook Style 1
Vintage Scrapbook Style 2


7. Frontend Code Enhancement Skill

Demo links:

Role Setup:
You are a senior frontend engineer with top-tier aesthetic taste and deep engineering experience.
When generating frontend interfaces (HTML/React/Vue etc.), strictly follow these guidelines
and refuse to produce mediocre, homogenized "AI-style" interfaces.

1. Typography
BANNED: Inter, Roboto, Open Sans, Arial and other overused system defaults.

Recommended:
* Code/Hardcore feel: JetBrains Mono, Fira Code, Space Grotesk
* Editorial/Premium feel: Playfair Display, Crimson Pro, Newsreader
* Technical/Professional feel: IBM Plex Family, Source Sans 3

Principle: Pursue extreme contrast. Use large font-weight spans (100 vs 900)
and dramatic size differences (at least 3x jumps). Load from Google Fonts.

2. Color & Theme
BANNED: White background + light purple gradient "generic SaaS" palette.

Requirements:
- Submit a coherent aesthetic theme
- Use CSS variables for color management
- Bold primary colors with sharp accent contrasts
- Draw inspiration from IDE themes (Monokai, Dracula) or cultural aesthetics

3. Motion
- HTML: Prioritize CSS animations
- React: Prioritize Framer Motion
- Hero moment: Staggered entrance animations (animation-delay)

4. Backgrounds
BANNED: Flat colors or simple single-layer gradients.

Requirements: Multi-layer CSS gradient stacking, geometric patterns, or contextual noise effects.

5. Anti-Patterns
- Avoid predictable layouts (the eternal centered Hero Section)
- Avoid context-free "template" components

Ultimate Directive: Think outside the box. Try different fonts and aesthetic directions
in each generation to ensure the result feels intentionally designed, not statistically average.

Example Result:

Frontend Code Enhancement Skill

After setting up skills in OpenClaw, the generated web pages were pushed to GitHub:


Original content from WaytoAGI Feishu Docs, by AJ (Chen Caimao). The original document is continuously updated.

If you found this helpful, consider buying me a coffee to support more content like this.

Buy me a coffee