hraness

saved

Ian Xiaohei illustrations

by Ian (伊恩)GitHub

gist

Ian (伊恩) publishes a Codex skill that turns judgments, processes, states, and metaphors in Chinese articles into 16:9 white-background hand-drawn figures. Xiaohei, a black stick figure with white-dot eyes, must perform the pictured action rather than decorate the frame. The skill is not a generic illustration prompt or a PPT infographic: it extracts one cognitive anchor per shot, invents a low-tech metaphor, and keeps Chinese annotation short. Example images calibrate style; agents should reinvent the composition from the current article.

ideas

  • Draw the cognitive action, not a decoration. The skill exists to picture one judgment, process, state, or metaphor from the article.
  • Xiaohei must do the work. If removing the character leaves the diagram intact, he is only a sticker.
  • One structure per frame. Do not dump a full article, course page, or PPT architecture into a single 16:9 shot.
  • Examples calibrate style, not composition. Reinvent the metaphor from the current text instead of copying old objects and layouts.
  • Short Chinese labels are more reliable. If generated text is garbled, cut annotation words and regenerate.

quotes

它不是通用插画 prompt,也不是 PPT 信息图模板。

Ian (伊恩), distinguishing the skill from generic illustration prompts and PPT templates.

小黑不是吉祥物,不是贴纸,也不是站在角落里的装饰物,而是正在认真参与系统运转的荒诞工作者。

Ian (伊恩), defining Xiaohei as a worker inside the pictured system.

这些图片是风格校准样例,不是构图模板。

Ian (伊恩), warning against copying example compositions.

每张图只讲一个核心结构,不要把文章做成说明书。

Ian (伊恩), limiting each figure to one structure.