Ian Xiaohei Scenes skill
A Codex skill that draws 16:9 scenes with the Xiaohei character among real objects for Chinese articles
- 544 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Reads the article text and saves generated PNGs into the working folder
- Sends frame descriptions to an external image model over the network
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/ian-xiaohei-scenesDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Reads the article text and saves generated PNGs into the working folder
- Sends frame descriptions to an external image model over the network
This is third-party code. Review the repository files before installing.
What it does
The skill turns a situation from a Chinese article, post, project retrospective or career story into a picture: a small black character, Xiaohei, performs one physical action among photorealistic objects on a clean white background. From the text the agent pulls out the reader's position, the core conflict and the object that carries it best. It then invents the action, picks two to four short Chinese labels and writes a separate prompt for an image model. Standard mode produces a 16:9 frame for an in-text illustration. Before generating, the agent picks one of six reference frames in assets/examples, from a message avalanche to resume screening. The reference sets proportions, whitespace and label density. The agent invents the objects, pose and composition anew. Easter-egg mode builds a long horizontal strip of five to eight object nodes along one hand-drawn line. It suits the story of a career, a project or a product's evolution. Facts in those nodes come only from the user's text, and anything that cannot be confirmed the agent marks as pending or replaces with a general word.
Who it is for. For authors of Chinese articles, posts and project write-ups who work in Codex and want one recognizable illustration style.
Good fit when
- You need three or four 16:9 illustrations for an article on work stress, message overload or review rework
- You want only a shot list before generating: tell the agent not to generate yet, and it returns a plan with labels and a rationale for each frame
- You need a long story strip for a project retrospective or a career story where every node rests on an object and the character's action
- You run a blog or channel in a Chinese-speaking audience and want one visual language across posts
Not a fit when
- You need a commercial key visual, a brand poster or a polished flat illustration
- You need a flowchart, architecture diagram or course slide: the author points to the Ian Xiaohei Illustrations skill for hand-drawn diagrams
- You need interface screenshots, chat windows or dashboards
- You need an editable vector source: the result is a raster PNG
Example request
Use $ian-xiaohei-scenes: do not generate yet, propose five frames for an article on IT burnout and name the reference you take for eachLimitations
- The skill gives Codex instructions, and an image model connected in your Codex draws the picture.
- The skill's text and the labels on images are written for Chinese. The shorter the label, the fewer the errors. The model can still distort characters or add a stray caption, so check finished frames by eye.
- The reference frames must not be traced or redrawn. If a new picture looks like an old one with different text, the skill's checklist counts it as a failure.
- Only the nested ian-xiaohei-scenes folder goes into Codex. The root README, NOTICE and examples serve the GitHub page.
How to disable. Delete the ian-xiaohei-scenes folder from the Codex skills directory, which is ~/.codex/skills by default. The cloned repository can be deleted separately.
FAQ
How do I install Ian Xiaohei Scenes in Codex?
Clone the repository, create the skills directory with mkdir -p and copy the nested ian-xiaohei-scenes folder into it with cp -R. The new skill appears after Codex restarts.
How is Ian Xiaohei Scenes different from Ian Xiaohei Illustrations?
Ian Xiaohei Illustrations, the first version, draws hand-drawn explanatory diagrams on white to unpack a method, process or structure. Scenes, version 2.0, places the character in a scene of real objects, as in a white photo studio, and fits emotions, work pressure and project stories.
How many labels and actions does one picture allow?
The skill requires one physical action, one main object or a compact group, at most one or two small props and at most four labels, preferably three. To cut clutter, the agent trims the object list first and generates afterwards.
Can I get just an illustration plan without generating?
Yes. Ask the agent not to generate and to produce a shot list first. For each frame it describes the reference, topic, the feeling the reader recognizes, the action, the main object, the character's behavior and the labels.
How do I choose between standard mode and easter-egg mode?
For one idea in the text take standard mode: a 16:9 frame with one action. For a story about a path, a project or a product's development ask for easter-egg mode, an ultra-wide strip with five to eight nodes, close to 2.6:1–3:1 in proportions.
Discussion
No questions or reviews yet.
Related
SkillsA Codex skill that turns a Chinese article into a series of hand-drawn 16:9 illustrations on white background with the Xiaohei character
A Codex skill that turns articles and notes into a set of page images in a Chinese handdrawn technical explanation style
A skill that turns photos of buildings, streets and local culture into art-directed editorial posters via an image model
A skill that turns your photo or description into a hand-drawn character and uses it for diagrams, knowledge cards and slides
A skill that takes an agent from a generated picture to a game asset: sprites, video-to-frames animation and pre-release checks
A set of skills for Claude Code and Codex that assemble detailed image prompts from an element library and composition schemes
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Repository stars | Main difference |
|---|---|---|---|
| Ian Xiaohei Scenes (this entry) | no | 544 | |
| Ian Xiaohei Illustrations | no | 12,981 | the author's first version, hand-drawn explanatory diagrams on white with no scenes of objects |
| Ian Handdrawn PPT | no | 1,497 | hand-drawn pages in a technical presentation style, no character in a scene |
README in short
The README explains the formula of character plus object plus action plus short label plus whitespace and compares versions 1.0 and 2.0. The examples include six standard-mode references and one strip reference, followed by installation, sample requests, workflow stages and the folder layout.
README badge
Original SKILL.md
--- name: ian-xiaohei-scenes description: | 生成“小黑 + 真实物件 + 物理动作 + 留白叙事”的中文配图。默认输出 16:9 正文配图，用于互联网打工人共鸣图、AI 时代职场焦虑、程序员/产品/创作者处境和正文观点隐喻图；遇到“彩蛋模式 / 长卷故事图 / 超横版 / 个人经历 / 项目复盘 / 产品演化 / 成长路径”时，输出小黑长卷故事图。标准模式默认 #FFFFFF 纯白背景；彩蛋长卷模式使用高级近白背景和一条真实物件人生线。样例是高质量模板母版和出图质量标尺；必须对齐其比例、留白、动作清晰度和叙事关系，但不能复刻其物件组合、空间拓扑、小黑姿态或标签位置。 when-to-use: | 用户需要为中文文章、帖子、教程、案例、项目复盘或个人经历生成“小黑 + 真实物件 + 物理动作”的高质量配图时触发。 触发词：小黑实物场景图、小黑 2.0、实物图、真实物件、小黑 + 物件、打工人共鸣图、正文配图、彩蛋模式、长卷故事图、超横版、个人经历、项目复盘、产品演化、成长路径。 输入状态：可以是一段正文、一个主题、一组节点、一个项目/经历说明，或用户给定的参考母版。 allowed-tools: - Read - Write - Edit - Bash context: fork --- # Ian 小黑实物场景图 ## 核心定位 把文章观点、用户处境、个人经历、项目过程、产品演化和内容资产，转译成一张可传播的“真实物件小现场”或“真实物件人生线”： ```text 小黑 + 真实物件 + 物理动作 + 短中文标签 + 留白叙事 ``` 这不是 1.0 的纯手绘解释图，也不是 PPT 信息图。2.0 的目标是让读者先看到一个真实、轻、怪的现场，再在 1 秒内意识到“这说的就是我”。 本 Skill 有两种输出规格： - **标准模式**：16:9 正文配图，一个真实物件小现场，默认纯白 `#FFFFFF` 背景。 - **彩蛋模式**：小黑长卷故事图，一条超横版真实物件人生线，用于个人经历、项目复盘、产品演化、成长路径和内容资产演进。它以 `assets/examples/07-long-scroll-story-master.png` 为核心母版，使用高级近白背景，不强制死白。 ## 和 1.0 的区别 - 1.0：更像白板上的手绘逻辑图，适合拆文章、拆流程、拆方法。 - 2.0：更像白色摄影棚里的真实物件现场，适合表达处境、情绪、项目故事、产品故事和正文观点隐喻。 如果用户要求“纯手绘小黑解释图 / 白板风流程图 / 手绘方法论图”，优先用 `$ian-xiaohei-illustrations`。如果用户要求“真实物件 / 实物图 / 小黑 + 物件 / 打工人共鸣 / 更有生活质感 / 小黑 2.0”，用本 Skill。 ## Before Starting 按任务需要读取，不要一次全部塞进上下文： - `references/style-dna.md`：2.0 视觉 DNA、比例、留白、颜色、真实物件规则。