Claude-skill-registry ACOS Visual Generator

Generate research-grounded visuals using the InfoGenius pipeline. Use when creating infographics, diagrams, educational visuals, or any image that benefits from factual accuracy. Supports 8 visual styles (3D, technical, minimalist, photorealistic, futuristic, vintage, cartoon, standard) and 4 audience levels.

install

source · Clone the upstream repo

git clone https://github.com/majiayu000/claude-skill-registry

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/acos-visual-gen" ~/.claude/skills/majiayu000-claude-skill-registry-acos-visual-generator && rm -rf "$T"

manifest: skills/data/acos-visual-gen/SKILL.md

source content

ACOS Visual Generator

What This Skill Does

Creates research-grounded images by:

Researching the topic via web search
Extracting key facts for accuracy
Constructing a detailed image prompt
Generating via Nano Banana MCP (Gemini 2.5 Flash Image)

Prerequisites

Nano Banana MCP server configured
WebSearch tool available

Quick Start

User: "Create a 3D infographic about quantum computing"

Process:
1. WebSearch → Extract key facts
2. Construct prompt with style instructions
3. mcp__nanobanana__generate_image()
4. Return image with sources

Visual Styles

Style	Description
`standard`	Clean scientific illustration, modern, professional
`minimalist`	Bauhaus, flat vector, 2-3 colors, negative space
`photorealistic`	Cinematic lighting, 8K, detailed textures
`3d`	Isometric render, claymorphism, studio lighting
`technical`	Da Vinci notebook, ink sketch, annotations
`futuristic`	Cyberpunk HUD, neon, holographic
`vintage`	19th century lithograph, sepia, engraving
`cartoon`	Educational comic, vibrant, cel-shaded

Audience Levels

Level	Target
`elementary`	Ages 6-10, bright, simple, fun icons
`highschool`	Standard textbook, clean, accurate
`college`	Academic journal, high detail, data-rich
`expert`	Technical schematic, extremely dense

Step-by-Step Process

Step 1: Research

WebSearch("{topic} facts key information 2026")

Extract 3-5 verifiable facts.

Step 2: Construct Prompt

Create a {aspect_ratio} {style_description} infographic about {topic}.

VISUAL STYLE: {style_instruction}
AUDIENCE: {audience_instruction}

INCLUDE:
- {fact_1_visualized}
- {fact_2_visualized}
- {fact_3_visualized}

COMPOSITION:
- Clear layout with visual hierarchy
- Text should be large and legible
- Include labels and annotations

Step 3: Generate

mcp__nanobanana__generate_image({
  prompt: "{constructed_prompt}",
  aspect_ratio: "16:9",
  model_tier: "pro",
  enable_grounding: true,
  thinking_level: "high",
  resolution: "high"
})

Step 4: Present Results

Return:

Generated image
Research sources
Key facts included
Modification options

Style Instructions (Copy These)

Standard:

High-quality digital scientific illustration. Clean, modern, highly detailed. Professional color palette.

3D Isometric:

3D Isometric Render. Claymorphism or high-gloss plastic texture, studio lighting, soft shadows, looks like a physical model.

Technical:

Da Vinci Notebook style. Ink on parchment sketch, handwritten annotation style, rough but accurate lines, technical precision.

Futuristic:

Cyberpunk HUD. Glowing neon blue/cyan lines on dark background, holographic data visualization, 3D wireframes.

Example Usage

User: "Visualize how neural networks learn"

Research: WebSearch finds key facts about backpropagation, gradient descent, layers, activation functions.

Prompt:

Create a 16:9 3D isometric infographic about how neural networks learn.

VISUAL STYLE: 3D Isometric Render with claymorphism aesthetic...

INCLUDE:
- Input layer receiving data
- Hidden layers processing with weights
- Backpropagation arrows showing error correction
- Output layer with predictions

Generate: Call Nano Banana MCP with grounding enabled.

Integration

This skill integrates with:

```
/infogenius
```
command in ACOS
Nano Banana MCP for image generation
WebSearch for factual grounding

Part of the Agentic Creator OS visual generation pipeline.