---
name: Oil Visual
slug: oil-visual
category: AI Engineering
description: "Create consistent manga-ink illustrations in two modes: explanatory images with embedded labels, or transparent character scenes for layouts. Ideal for explaining concepts, workflows, comparisons, and hero artwork."
github: "https://github.com/oil-oil/oil-visual"
language: Python
stars: 66
forks: 13
install: "npx degit https://github.com/oil-oil/oil-visual ~/.claude/skills/oil-visual"
installs_to: ~/.claude/skills/oil-visual
source_path: SKILL.md
collection_size: 1
category_size: 2451
added: 2026-08-11T07:21:11.323Z
last_synced: 2026-08-11T07:21:11.323Z
canonical_url: "https://dirskills.com/skills/oil-visual"
---

# Oil Visual

Create consistent manga-ink illustrations in two modes: explanatory images with embedded labels, or transparent character scenes for layouts. Ideal for explaining concepts, workflows, comparisons, and hero artwork.

**Install:**

```bash
npx degit https://github.com/oil-oil/oil-visual ~/.claude/skills/oil-visual
```

## README

<p align="center">
  <img src="./assets/readme/hero.svg" width="100%" alt="oil-visual：漫画墨线、半调网点、小人和黄色边牧组成的配图风格">
</p>

`oil-visual` 是一套给 Codex 使用的配图规范。它把角色、线条、网点、颜色、场景和图片文字统一在一套规则里，让不同主题的图保持同一种视觉语言。

适合用来画概念解释、机制流程、对比关系和角色场景。

## 配图风格

- 用粗细分明的黑色墨线画轮廓和细节。
- 用圆形半调网点表现灰面和阴影，不使用柔和渐变。
- 主角是戴圆框眼镜的极简小人，旁边保留一只胖胖的暖黄色边牧。
- 场景使用米白色纸张质感，只保留少量真正帮助理解的环境物件。
- 暖黄色是固定颜色；每张图最多再使用两种语义色。
- 图片需要文字时，使用短而准确的无衬线标签，并直接放在对应物体上。

## 示例

<p align="center">
  <img src="./examples/agent-workflow-square.png" width="32%" alt="主 Agent 分配任务并统一验收">
  <img src="./examples/explainer-cost-comparison.png" width="32%" alt="分工清楚与重复处理的成本对比">
  <img src="./examples/feedback-loop.png" width="32%" alt="输入、处理、结果与反馈组成的循环">
</p>

<p align="center">
  <img src="./examples/pipeline-bottleneck.png" width="32%" alt="并行处理疏通任务瓶颈">
  <img src="./examples/from-complex-to-clear.png" width="32%" alt="把复杂输入整理成清楚结果">
</p>

同一套小人、边牧、墨线、网点和语义色，可以表现机制、对比、循环、瓶颈和信息整理。

## 安装

```bash
git clone https://github.com/oil-oil/oil-visual.git ~/.codex/skills/oil-visual
```

重启 Codex，然后在对话里点名使用：

```text
Use $oil-visual to explain how the main Agent assigns work to subagents.
Use short, exact Chinese labels inside the image.
```

```text
Use $oil-visual to draw the glasses stick figure reviewing a blueprint,
with the warm-yellow Border Collie sitting beside the desk.
```

完整规则、提示词结构和交付检查都写在 [`SKILL.md`](./SKILL.md) 里。

## License

[MIT](./LICENSE)