Guide · Agent skill
How to use the baoyu-comic agent skill: knowledge comics
Updated · written with AI from the project's own documentation and checked against our catalogue data
baoyu-comic is an agent skill that lets an AI agent plan and draw an original educational comic from a topic, a biography or a tutorial. It is written for people who have never used agent skills, so this guide starts with what a skill is and then covers installation, triggering, requirements, and safety. By the end you will know how to add the skill, what to ask for, and what you need in place before the first page is drawn.
Knowledge Comic CreatorJimLiu/baoyu-skills · MIT Download the skill Quick start
- 1Download and unzip. Download the skill ZIP from this page and unzip it on your computer.
- 2Place the folder. Move the baoyu-comic folder into ~/.claude/skills/ or a project's .claude/skills/ folder, or upload it under Settings → Capabilities → Skills in the Claude apps.
- 3Set up an image backend. Make sure one supported backend is available, such as the codex CLI with an active login or baoyu-image-gen, and that bun or npx is installed.
- 4Ask for a comic. Describe the topic in plain language and include the words 知识漫画, biography comic, or tutorial comic if they fit.
- 5Answer style and tone questions. When the agent asks, choose an art style such as manga, ink brush, chalk or minimalist, and pick a tone.
- 6Review prompts and pages. Check the prompt files saved under prompts/ and the generated pages, then ask for regeneration if any text is wrong.
What the skill makes the agent do
Once installed, the skill gives the agent a workflow for making a multi-panel comic. The agent analyses your source material, plans the story and panel layouts, and writes a prompt file for each image. It asks you about art style and tone, then calls an image generation tool to draw the pages.
The skill includes reference notes on several art styles, including manga, ligne claire, ink brush, chalk, minimalist, and realistic. It also includes a character template, preference settings, and a watermark guide.
- Analyses a topic, biography or tutorial before drawing
- Plans panel layouts and writes one prompt file per image under prompts/
- Asks about style and tone where needed
- Generates images in batches using the backend you have available
Install the baoyu-comic skill
Download the ZIP for the skill from this page and unzip it. Then place the folder in one of two locations, depending on how you work.
For a personal setup, put it in ~/.claude/skills/. For a single project, put it in that project's .claude/skills/ folder. If you use the Claude apps, you can instead upload the skill under Settings → Capabilities → Skills.
- Personal: ~/.claude/skills/
- Per project: .claude/skills/ inside your project folder
- Claude apps: Settings → Capabilities → Skills, then upload
Know what you need before you start
The skill does not draw images itself. It needs an image generation backend. The skill checks for one of the following, in this order of preference when several are available: Codex imagegen, the codex CLI with an active login, Cursor GenerateImage, or baoyu-image-gen. You can also name a backend in your request.
The skill metadata lists bun or npx as required tools, and the skill uses network access and runs scripts.
- An image generation backend: Codex imagegen, codex CLI with login, Cursor GenerateImage, or baoyu-image-gen
- bun or npx available on your system
- Network access, because the image backend calls out to a service
Trigger the skill with a natural request
You do not need special syntax. Ask for a comic in plain language. The skill's description lists phrases such as 知识漫画, 教育漫画, biography comic, tutorial comic, and Logicomix-style comic, so these are good starting points.
Be specific about the subject, the audience, and the length if you know them. The agent will ask about style and tone when it needs your choice.
- "Make a biography comic about Marie Curie for high school students."
- "Create a tutorial comic explaining how a binary search works."
- "Make a 知识漫画 about the water cycle in ink brush style."
Check the results and the limits
Output quality depends on the image backend you connect. The skill forbids painting over generated text, so if a speech bubble or label is wrong, the agent should regenerate the image from a corrected prompt, redraw with less text, or ask you which candidate to keep. Do not expect exact on-image text every time.
The skill is for original work. Do not use it to reproduce existing copyrighted comics or characters. You are responsible for rights and attribution in what you publish.
- Review every page for text accuracy before sharing
- Use it for original content, not copies of existing comics
Safety notes: scripts and network
This skill runs scripts and uses the network. Review the files in the skill folder before you install it, and check which backend and tools your agent is allowed to run. Keep the skill's own folder structure intact so the reference files are found.
- Runs scripts: yes
- Uses network: yes
- The skill's SKILL.md and references/ folder describe the workflow and style notes
What people use it for
- Explaining a science topic to students as a short visual lesson
- Turning a historical figure's life into a biography comic
- Writing a step-by-step tutorial as a comic for a workshop or blog
- Creating a consistent cast of characters across several comic pages
Frequently asked questions
How do I install the baoyu-comic skill?
Download the skill ZIP from this page and unzip it. Place the folder in ~/.claude/skills/ for personal use or in a project's .claude/skills/ folder, or upload it in the Claude apps under Settings → Capabilities → Skills.
Is baoyu-comic free to use?
The skill is published under the MIT licence. It does not provide an image generator, so any cost depends on the backend you connect, such as Codex imagegen or baoyu-image-gen.
What do I need to run baoyu-comic?
You need an image generation backend: Codex imagegen, the codex CLI with login, Cursor GenerateImage, or baoyu-image-gen. You also need bun or npx on your system.
Can baoyu-comic make a comic from an existing copyrighted comic?
No. The skill is for original work. You are responsible for rights and attribution for anything you create or publish with it.
Why does the text in my comic look wrong?
The skill does not paint over generated text, because that is forbidden in its rules. Text errors are usually fixed by regenerating the page from a corrected prompt or by drawing with less on-image text.
How is baoyu-comic different from a plain image generator?
A plain image generator draws one picture from one prompt. This skill plans a multi-panel story, writes a prompt file for each image, asks about style and tone, and then generates the pages in batches.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use baoyu-comic if you already have an image generation backend and want original educational or biography comics with consistent characters and layouts. Pick a different tool if you need a finished comic with no setup, or if you need exact on-image text every time.
Download the skill