Back to directory

guizang-material-illustration

A Claude Code and Codex agent skill that turns articles, chart screenshots and reports into labeled illustrations, redrawn charts and reference-based diagrams in a consistent style.

guizang-material-illustration

This skill focuses on producing the one central image that an article, slide deck, social card or report needs to make an idea clear, rather than a decorative graphic that looks good but explains nothing. It covers three jobs: explainer diagrams that turn an abstract concept, process, mechanism or system relationship into a labeled illustration; chart beautification, which extracts the type, title, data, axes and conclusion from a rough chart screenshot and redraws it more legibly without copying the original layout; and reference-assisted illustration for unfamiliar concepts, brands, models or historical objects, which looks up supporting facts and reference images before drawing.

It is meant to be paired with a separate social-card skill that handles a card's title, body text and layout, while this skill supplies the central illustration. Output uses a consistent three-dimensional illustration style with a default blue theme, alternate accent colors, and in-image labels rather than plain decoration, and a QA pass checks labels, data accuracy, cropping and readability at social-card sizes.

It installs through a skills package-manager command or by cloning the repository into a Claude Code or Codex skills directory.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.