Ian Xiaohei Illustrations
A Codex Skill that guides AI agents to draw white-background hand-drawn Chinese article illustrations.
- Stars
- 13k
- Forks
- 1.5k
- Licence
- MIT
- Language
- —
- Last update
- 10 days ago
- Type
- AI model / agent
What it is
This is a Codex Skill, a set of instructions that tells an AI agent how to make 16:9 hand-drawn illustrations for Chinese articles, posts, and blogs. It follows a fixed visual style: white background, black line art, a small black character called Xiaohei, and a few red, orange, or blue Chinese handwritten notes. It is meant for article body images, not general illustration or slide-style infographics.
- Plan a shot list of 4-8 illustrations for a Chinese article
- Generate a 16:9 white-background hand-drawn image for one concept
- Edit an existing generated image to remove a stray title or error text
Great for
- Chinese-language writers who need body illustrations for articles or posts
- Knowledge and methodology content creators who want abstract ideas drawn as metaphors
- People using Codex who want a consistent visual style across many images
Not ideal for
- Anyone wanting commercial illustration, brand artwork, or polished flat design, since the style is intentionally sketchy and quirky
- Anyone needing PPT-style flowcharts or architecture diagrams, since the README says it is not an infographic template
- Anyone needing editable vector source files, since the default output is PNG only
How to get it
- 1
Clone and build from source
You'll need: Codex with skills support · Git
Clone the repository
git clone https://github.com/helloianneo/ian-xiaohei-illustrations.gitCopy the skill into the Codex skills directory
mkdir -p "${CODEX_HOME:-$HOME/.codex}/skills" && cp -R ./ian-xiaohei-illustrations "${CODEX_HOME:-$HOME/.codex}/skills/" - 2
Run it
- Clone the repository and copy the ian-xiaohei-illustrations subfolder into your Codex skills directory
- In Codex, invoke the skill by name with your article text, for example: Use $ian-xiaohei-illustrations to plan 5 illustrations for this article
- Review the shot list before generating images
- Generate the images and check them for typos and unwanted text
- 3
Get the most out of it
- Ask for shot-list planning first, without image generation, to choose which points get illustrated
- Keep Chinese labels short, since the README says shorter text renders more reliably
- Check that Xiaohei performs the core action; the README says if the image still works without it, the character is too decorative
- Use the example images only to calibrate style, not to copy composition
Downloads
Strengths
- Clear, narrow visual style that the README describes in concrete terms
- Includes guidance on planning shots before generating images
- Installs by copying a folder into the Codex skills directory
Trade-offs
- Chinese text in generated images can contain typos, so the README advises short labels and regeneration
- Output depends on an external image model, which the README says may drift in style or add unwanted titles
- The README does not specify which image model is required or how it is accessed
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| ian-xiaohei-illustrations (this) | 13k | AI model / agent | intermediate | 10 days ago | Chinese-language writers who need body illustrations for articles or posts |
| excalidraw | 135k | App | beginner | 4 days ago | Teams sketching diagrams or wireframes together in a shared canvas |
| mermaid | 91k | Dev tool | beginner | 4 days ago | Developers who want diagrams stored as text alongside their code and docs |
| hermes-agent | 260k | AI model / agent | intermediate | 4 days ago | Developers who want a personal agent they can host on their own server or machine |
| career-ops | 79.4k | AI model / agent | intermediate | 4 days ago | Job seekers who apply to many listings and want to filter them before spending time on each one |
| ppt-master | 72.4k | AI model / agent | intermediate | 7 days ago | People who want an editable starting deck from existing documents rather than a filled-in template |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.