Skip to main content
skill CX Codex

reimagine-it-infographic — Claude Code skill: setup and features

Generates a statistical data poster directly from facts found in your source file. It avoids generic dashboard templates and pie charts.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add Kayforkind/reimagine-it --skill infographic -a claude-code
Open on GitHub

What reimagine-it-infographic does

This skill turns source text into a statistical poster without relying on stock dashboard templates. It extracts anchors like dates and quantities directly from your text. Visual encodings follow Cleveland-McGill perceptual rankings, prioritizing position on a shared scale over angles or area.

Before rendering, the skill routes data into structures such as sequences, hierarchies, or relations. Countable items turn into ISOTYPE pictograms, where one symbol represents a fixed unit count. Color palettes derive directly from the source subject and stay under five tones.

Circular gauges, pie charts, 3D effects, and fabricated metrics are prohibited. Every number must match the input. A lossless data table must accompany the finished visual.

What it helps with

  • turn raw report text into a structured statistical poster
  • represent countable units using ISOTYPE pictogram arrays
  • compare metrics along common scales without visual distortion
  • supply a complete lossless verification table alongside charts

Example prompts for the agent

  • /reimagine-it infographic for report.md
  • Build a statistical poster using the data in metrics.csv
  • Create an infographic from this document without pie charts

How to install reimagine-it-infographic

1Skills CLI (Vercel)

npx skills add Kayforkind/reimagine-it --skill infographic -a claude-code

2Manually: put the files in place

~/.claude/skills/infographic/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/infographic/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about reimagine-it-infographic

How to install reimagine-it-infographic in Claude Code?

Run npx skills add Kayforkind/reimagine-it --skill infographic -a claude-code. The skill file lands in ~/.claude/skills/infographic/SKILL.md.

What does reimagine-it-infographic do?

It reads facts inside a file and lays out an editorial statistical poster. The skill extracts palette, structure, and glyphs directly from source context.

Why does this skill reject pie charts and gauges?

The skill follows Cleveland-McGill rankings where angles distort human perception. Comparisons rely on position along a common scale instead.

Source text

SKILL.md · MIT
---
name: reimagine-it-infographic
description: >-
  Content-derived infographic poster — reads the source file and builds a
  statistical poster from the facts already in it. Not a dashboard, not a CV,
  not an AntV template. Uses common-scale encodings (Cleveland–McGill floor),
  ISOTYPE unit counts (Neurath), custom data glyphs (Lupi/Fragapane),
  and a lossless data table. Palette, glyphs, and composition are derived
  from the source's own content. No pies, donuts, gauges, 3D charts, or
  fabricated KPIs. Part of the reimagine-it Content-Derived Design suite.
license: MIT
metadata:
  author: Kayforkind
  version: "2.15.0"
  parent: reimagine-it
  hosts:
    - claude-code
    - cursor
    - codex
    - copilot
    - gemini-cli
keywords:
  - infographic
  - data-visualization
  - statistical-poster
  - isotype
  - content-derived-design
  - cleveland-mcgill
  - tufte
  - data-humanism
category: Design
capabilities:
  - infographic-creation
  - data-visualization
  - statistical-poster-design
  - isotype-generation
  - custom-glyph-creation
trigger_phrases:
  - /reimagine-it infographic
  - make an infographic
  - create a data poster
  - statistical poster
  - visualize this data
  - info graphic
---

# /reimagine-it infographic

**Parent:** [../SKILL.md](../SKILL.md) — the full Content-Derived Design engine.
**Research pack:** [../references/research/infographic-craft.md](../references/research/infographic-craft.md) — 15-source infographic method (Cleveland–McGill, Tufte, Bertin, ISOTYPE, Minard/Snow, Cairo, Lupi, FT Visual Vocabulary, USWDS, WCAG charts, InfoAlign layouts).
**Domain pack:** [../references/domains/infographic.md](../references/domains/infographic.md) — the infographic domain non-negotiables.

Builds a **statistical poster of facts already in the source** — not an ops dashboard, not a CV, not an AntV template clone.

## Hard non-negotiables

1. **Cleveland–McGill perceptual ranking.** Position along a common scale > length > angle > area > volume > color saturation. Encode the most important comparisons as positions, not pies.
2. **ISOTYPE unit counts** when the data is discrete countable things (Neurath: one pictogram = N units, repeated, not scaled).
3. **Custom glyphs from the source.** Lupi/Fragapane pattern — draw a visual alphabet from the data itself (leaves, braids, snakes, whatever the content demands), not generic bar/pie/donut.
4. **Lossless data table.** Every chart has a companion table; the infographic is a supplement to the data, not a replacement.
5. **No pies, donuts, gauges, or 3D charts.** Pie charts rank last in Cleveland–McGill; donuts are pies with the center punched out (worse); 3D distorts perception.
6. **No fabricated KPIs.** Every number on the poster comes from the source, not extrapolated or invented.

## Structure router

Before drawing, sniff the source and lock a **structure**:

| Structure | When | InfoAlign scenery |
|-----------|------|-------------------|
| `list` | Unordered peers | grid, star |
| `sequence` | Steps / dated beats | portrait, landscape |
| `hierarchy` | Tree / nesting | portrait |
| `compare` | Two+ groups | portrait-grid |
| `values` | Magnitudes | portrait-grid |
| `relation` | Nodes + edges the source names | spiral |

**Do not default to gold Texas (sequence+compare on portrait-grid).** Derive from THIS source.

## Procedure

1. Load [../references/research/infographic-craft.md](../references/research/infographic-craft.md) and [../references/domains/infographic.md](../references/domains/infographic.md).
2. Extract anchors from the source (3–5 concrete nouns, dates, numbers).
3. Route the structure (list/sequence/hierarchy/compare/values/relation).
4. Pick InfoAlign scenery (grid/star/portrait/landscape/portrait-grid/spiral).
5. Derive palette from source content (≤ 5 colors).
6. Build a custom visual alphabet from one datapoint.
7. Encode comparisons using common-scale positions (not pies).
8. Include a lossless data table below the poster.
9. Verify: every number traceable to source, no fabricated KPIs, no AntV template chrome.

## Must not

- Clone gold Texas infographic's scenery or poster chrome onto a different source
- Use a pie chart, donut, gauge, or 3D chart
- Invent statistics the source does not contain
- Use AntV/Looker/Datawrapper templates
- Skip the structure router and default to portrait-grid
- Ship without a lossless data table

© Kayforkind · MIT · Kayforkind/reimagine-it

Similar skills

Everything for Design →

From the same set7

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor