infographic-creator

Turning blocks of text into clear, memorable graphics is the job of this skill, which drives AntV Infographic to render visual infographics from supplied content. It treats an infographic as Information Structure plus Visual Expression, and expresses both through a custom indentation-based DSL rather than JSON or markup. Every specification opens with a line reading infographic followed by a template name, then optional data and theme blocks indented two spaces.

The DSL organizes content around three ideas: template (the structure that carries the information), data (title, description, and template-specific items), and theme (palette, font, and style). Key-value pairs are written as key space value and array entries are prefixed with a dash, while icons are named with keywords such as star fill. Choosing the right main data field matters: list templates use lists, sequence templates use sequences, compare templates use compares, chart templates use values, relation templates use nodes and relations, hierarchy-structure uses items, other hierarchy templates use a single root nested through children, and items serves as a fallback when uncertain. Binary compare and left-right hierarchy templates require exactly two root nodes.

A large catalog of templates spans chart, compare, hierarchy, list, relation, and sequence families - from chart-pie-donut-plain-text and compare-swot to sequence-timeline-simple and relation-dagre-flow-tb-badge-card. Selection guidance maps intent to family: strict processes and steps map to sequence templates (timelines, staircases, roadmaps, zigzags, pyramids), while opinion listings map to list templates. Themes can be customized with palettes and fonts, and theme.stylize applies built-in looks including rough for a hand-drawn effect, pattern fills, and linear or radial gradients. The output is pure DSL - no JSON, Markdown, or explanatory prose - so it can be streamed directly into the AntV Infographic renderer.