alterlab-infographics

Infographics

Overview

Infographics are visual representations of information, data, or knowledge designed to present complex content quickly and clearly. This skill uses Nano Banana Pro AI for infographic generation with Gemini 3.1 Pro quality review and Perplexity Sonar for research.

How it works:

• (Optional) Research phase: Gather accurate facts and statistics using Perplexity Sonar
• Describe your infographic in natural language
• Nano Banana Pro generates publication-quality infographics automatically
• Gemini 3.1 Pro reviews quality against document-type thresholds
• Smart iteration: Only regenerates if quality is below threshold
• Professional-ready output in minutes; no design skills required

Quality Thresholds by Document Type:

Document Type	Threshold	Description
marketing	8.5/10	Marketing materials - must be compelling
report	8.0/10	Business reports - professional quality
presentation	7.5/10	Slides, talks - clear and engaging
social	7.0/10	Social media content
internal	7.0/10	Internal use
draft	6.5/10	Working drafts
default	7.5/10	General purpose

When to Use This Skill

Use the infographics skill when:

• Presenting data or statistics in a visual format
• Creating timeline visualizations for project milestones or history
• Explaining processes, workflows, or step-by-step guides
• Comparing options, products, or concepts side-by-side
• Summarizing key points in an engaging visual format
• Creating geographic or map-based data visualizations
• Building hierarchical or organizational charts
• Designing social media content or marketing materials

Does NOT Trigger

Scenario	Use Instead
Technical flowchart, CONSORT/PRISMA diagram, biological pathway, circuit, or neural-network architecture	alterlab-scientific-schematics
Photo, illustration, concept art, or hero image with no data layout	alterlab-generate-image
Exact chart of a dataset (every value must match the data) for a paper or report	alterlab-matplotlib / alterlab-scientific-viz
Full research poster for a conference session	alterlab-latex-posters

Quick Start

Generate any infographic by describing it; the script handles generation, review, and iteration.

# Generate a list infographic (default threshold 7.5/10)
python scripts/generate_infographic.py \
  "5 benefits of regular exercise" \
  -o figures/exercise_benefits.png --type list

# Marketing (highest threshold: 8.5/10)
python scripts/generate_infographic.py \
  "Product features comparison" \
  -o figures/product_comparison.png --type comparison --doc-type marketing

# Corporate style + colorblind-safe palette
python scripts/generate_infographic.py \
  "Company milestones 2010-2025" \
  -o figures/timeline.png --type timeline --style corporate

# WITH RESEARCH for accurate, up-to-date data
python scripts/generate_infographic.py \
  "Global AI market size and growth projections" \
  -o figures/ai_market.png --type statistical --research

What happens behind the scenes:

1. (Optional) Research: Perplexity Sonar gathers accurate facts, statistics, and data
2. Generation 1: Nano Banana Pro creates the initial infographic
3. Review 1: Gemini 3.1 Pro evaluates quality against the document-type threshold
4. Decision: quality >= threshold → DONE (early stop); below threshold → improve prompt and regenerate
5. Repeat until quality meets threshold OR max iterations reached

Output: Versioned images plus a detailed JSON review log with quality scores, critiques, and early-stop info. With --research, a _research.json file also records the gathered facts and the source URLs returned by the search model.

Check the numbers before you publish. The image model draws every figure and label itself, so it can misspell words or alter digits even when the prompt was correct, and the review model scores design quality, not factual accuracy. Compare each statistic in the finished graphic against your data or the _research.json sources, and cite those sources wherever the infographic is used. If the review call fails, the script keeps the image and records "review_skipped": true with a null score in the log instead of inventing one.

Core Workflow

1. Pick the type (--type) — one of 10 presets (statistical, timeline, process, comparison, list, geographic, hierarchical, anatomical, resume, social). Full catalog with examples in references/type_catalog.md.
2. Pick a style and (optionally) palette — 8 industry --style presets, 3 colorblind-safe --palette presets. Tables in references/type_catalog.md.
3. Set --doc-type to choose the quality threshold (see table above).
4. Add --research when accurate sourced data matters.
5. Configure the API key (OPENROUTER_API_KEY) and run; review the JSON log and regenerate with a more specific prompt if needed.

Configuration

Set the OpenRouter API key in the environment, or in a .env file if python-dotenv is installed:

export OPENROUTER_API_KEY='your_api_key_here'   # https://openrouter.ai/keys

Default models (verified on OpenRouter 2026-09-23): google/gemini-3-pro-image (Nano Banana Pro) for generation, google/gemini-3.1-pro-preview for review, and perplexity/sonar-pro for research. Override the first two with ALTERLAB_IMAGE_MODEL / ALTERLAB_REVIEW_MODEL.

Data & privacy: prompts and research/infographic content are sent to a third-party API (OpenRouter) for generation and quality review. Avoid sending confidential, clinical, or unpublished material.

Reference Index

File	Contents
references/type_catalog.md	All 10 --type presets with runnable examples; --style and --palette tables
references/cli_and_workflow.md	Full CLI options, smart-refinement loop + Gemini review criteria + review-log schema, --research integration, prompt-engineering tips, troubleshooting, cross-skill integration, checklist
references/design_principles.md	Visual hierarchy, layout, typography
references/color_palettes.md	Full palette specifications
references/infographic_types.md	Extended layout templates for all types

Use this skill to create professional, accessible, and visually compelling infographics using Nano Banana Pro AI with intelligent Gemini 3.1 Pro quality review.