Baoyu Skills Tutorial: 20+ AI Agent Skills for Content Creation
Baoyu Skills Tutorial: 20+ AI Agent Skills for Content Creation
Xiaoxin Software AlternativesYou know the feeling: your AI agent writes great articles, but when it comes to cover images, infographics, translation, or video transcripts, you’re juggling between a dozen different tools. The content creation workflow gets fragmented.
Baoyu Skills fixes this. It’s a collection of 20+ portable AI agent skills by JimLiu (宝玉), covering content generation, AI images, and daily utilities. Install once with npx skills add, and your agent knows exactly which skill to call for each task.
What is Baoyu Skills? An open-source AI agent skill collection (MIT license) by JimLiu, with 20+ portable skill files covering Xiaohongshu image cards, infographics, SVG diagrams, cover images, slide decks, video transcripts, translation, Markdown processing, and WeChat publishing. One-command install for Claude Code, Codex, Cursor, and more.
Key Data:
- 🌟 GitHub Stars: Rapidly growing (Repository)
- 📦 Skills: 20+, organized into Content / AI Generation / Utility categories
- ⚖️ License: MIT License
- 🧠 Core Feature: Full content creation pipeline — from writing to visuals to publishing
- 📦 Core Feature: 10+ image generation providers (Google, OpenAI, Azure, DashScope, etc.)
- 📖 Companion book: “Illustrated Agent Skills”
Prerequisites
- Node.js installed (for
npxcommands) - An AI coding tool: Claude Code, Codex, Cursor, or similar
- API Keys (optional): needed only for image generation skills
Overview
This Baoyu Skills setup guide covers all 20+ skills in detail, organized into three categories: content skills, AI generation skills, and utility skills.
Step-by-Step Guide
Step 1: One-Command Install
1 | npx skills add jimliu/baoyu-skills |
💡 Tip: Don’t install all 20+ skills at once — each loaded skill occupies extra context. Install only what you need.
Step 2: Install in Your AI Tool
Claude Code: Run npx skills add jimliu/baoyu-skills. Claude auto-loads SKILL.md.
Codex (OpenAI): Copy needed skill directories to .agents/skills/ in your project.
OpenClaw: Install via ClawHub: clawhub install baoyu-image-gen
Universal: Tell your agent: “Please install the skills from github.com/JimLiu/baoyu-skills.”
Step 3: Choose Your Skills
Content Skills
baoyu-xhs-images: Xiaohongshu Image Cards
Generates 1-10 image cards from content with 12 visual styles × 6 layout options.
1 | /baoyu-xhs-images article.md --style notion --layout dense |
Styles: cute, fresh, warm, bold, minimal, retro, pop, notion, chalkboard, study-notes, screen-print, sketch-notes
Layouts: sparse (1-2 points), balanced (3-4), dense (5-8), list (4-7 items), comparison (side-by-side), flow (3-6 steps)
baoyu-infographic: Professional Infographics
20 layout types × 17 visual styles. Analyzes content and recommends the best combination.
1 | /baoyu-infographic content.md --layout pyramid --style corporate-memphis |
20 Layouts: bridge, circular-flow, comparison-table, do-dont, equation, feature-list, fishbone, funnel, grid-cards, iceberg, journey-path, layers-stack, mind-map, nested-circles, priority-quadrants, pyramid, scale-balance, timeline-horizontal, tree-hierarchy, venn
17 Styles: craft-handmade, claymation, kawaii, storybook-watercolor, chalkboard, cyberpunk-neon, bold-graphic, aged-academia, corporate-memphis, technical-schematic, origami, pixel-art, ui-wireframe, subway-map, ikea-manual, knolling, lego-brick
baoyu-diagram: SVG Diagram Generator
Generates publishable SVG diagrams — flowcharts, sequence diagrams, architecture diagrams, illustrative explanations, class diagrams. No image model called; Claude writes SVG code directly with auto dark mode support.
1 | /baoyu-diagram "How JWT authentication works" --type flowchart |
5 diagram types: flowchart, sequence, structural, illustrative, class
baoyu-cover-image: Cover Image Generator
Five-dimensional customization: Type × Palette × Rendering × Text × Mood. 11 palettes × 7 rendering styles = 77 unique effects.
1 | /baoyu-cover-image article.md --type conceptual --palette cool --rendering digital |
baoyu-slide-deck: Slide Deck Generator
Generates professional slide images from content. Creates a full outline first, then generates slides page by page.
1 | /baoyu-slide-deck article.md --style corporate --audience executives --slides 15 |
baoyu-image-gen: Universal Image Generation
Unified interface for 10+ providers: Google Gemini, OpenAI, Azure, DashScope, Z.AI, MiniMax, Replicate, Jimeng, Seedream, Agnes.
1 | /baoyu-image-gen "a cat writing code" --provider openai --model gpt-image-2 |
Utility Skills
baoyu-youtube-transcript: YouTube Subtitle Download
Downloads subtitles with multi-language support, translation, chapter segmentation, and speaker identification.
1 | /baoyu-youtube-transcript https://youtu.be/VIDEO_ID --languages zh,en --chapters --speakers |
baoyu-url-to-markdown: URL to Markdown
Fetches any URL via Chrome CDP and converts to Markdown. Saves HTML snapshots as backup.
baoyu-danger-x-to-markdown: Twitter to Markdown
Converts X (Twitter) content to Markdown. Supports tweet threads and X articles.
baoyu-translate: Three-Mode Translation
Quick (direct), Standard (analyze then translate), Refined (full editorial workflow with review and polish). Supports all major languages with audience-aware translation.
1 | /translate article.md --to zh-CN --mode refined --audience technical --style formal |
Features: auto-chunking for long documents (4000+ words), metaphor adaptation, cultural annotations, 9 style presets.
baoyu-format-markdown: Markdown Formatting
Adds frontmatter, headings, summaries, hierarchy, bold, lists, and code blocks to raw text. Auto-runs typographic scripts.
baoyu-markdown-to-html: Markdown to HTML
Converts to styled HTML with WeChat-compatible themes, code highlighting, and optional citation extraction.
baoyu-compress-image: Image Compression
Compresses images while maintaining quality.
baoyu-wechat-summary: WeChat Group Summary
Extracts topics, quotes, and statistics from WeChat group messages. Supports incremental mode, member profiles, and two output styles (normal and roasting).
baoyu-electron-extract: Electron App Source Extraction
Extracts resources and JavaScript from any Electron app’s app.asar. Restores original source trees when source maps are available.
Environment Configuration
Image generation skills need API Keys. Priority: CLI env vars → system env → project .baoyu-skills/.env → user ~/.baoyu-skills/.env.
Supported providers: Google Gemini, OpenAI, Azure OpenAI, DashScope, Z.AI, MiniMax, Replicate, Jimeng, Seedream, Agnes.
Custom Extensions
All skills support customization via EXTEND.md files at project or user level.
FAQ
Q: How do I install Baoyu Skills in Claude Code?
A: Run npx skills add jimliu/baoyu-skills. Claude Code auto-loads SKILL.md. Use /baoyu-xhs-images, /baoyu-cover-image etc. in conversation.
Q: Should I install all skills?
A: No. Each skill occupies extra context. Install only what you need. The minimum combo for WeChat publishing: baoyu-cover-image + baoyu-post-to-wechat.
Q: What’s the difference between baoyu-diagram and baoyu-infographic?
A: baoyu-diagram generates SVG code (vector, no image model). baoyu-infographic generates pixel-based infographics (20 layouts × 17 styles).
Q: What image generation providers are supported?
A: 10+ providers including Google, OpenAI, Azure, DashScope, Z.AI, MiniMax, Replicate, Jimeng, Seedream, Agnes.
Q: How do I update the skills?
A: Run /plugin in your agent, go to Marketplaces, select baoyu-skills, then Update marketplace.
Advanced Tips
- WeChat publishing pipeline: baoyu-format-markdown → baoyu-cover-image → baoyu-post-to-wechat
- Xiaohongshu content pipeline: Write article → baoyu-xhs-images generates card series → publish
- Infographic + diagram combo: baoyu-infographic for overview + baoyu-diagram for detailed flows
- EXTEND.md customization: Create custom brand palettes shared across your team
Conclusion
This Baoyu Skills tutorial and installation guide covers all 20+ skills from one-command setup to detailed usage. From content creation to daily efficiency tools, this skill collection upgrades your AI agent from “writes text” to “writes, draws, translates, and publishes.” Whether you’re making Xiaohongshu cards, professional infographics, or translating articles, this guide has you covered.
How to cite this article: This article is based on the Baoyu Skills official GitHub repository (verified 2026-06-24). All skill information verified against the latest version. Companion book: “Illustrated Agent Skills — AI Efficiency Practical Guide.”
