Collection of Customized Agent Skills

A curated collection of specialized Agent Skills designed for AI-assisted development workflows and productivity enhancement.

Overview

This directory contains custom Agent Skills developed to streamline repetitive tasks and enable consistent project scaffolding. Each skill is a self-contained module with its own documentation, scripts, and dependencies, ready to be invoked by AI assistants or executed directly.

---

Available Skills

1. astro-project-init

Initialize a new Astro project with Tailwind CSS and basic page scaffolding.

Description: Creates a standardized Astro project from scratch with the "Late Night Flight" dark theme, base components (Navbar, Footer, BaseLayout), and a minimal Home page for verification.

Use Cases:

• Create new Astro projects
• Set up Tailwind CSS with custom dark theme
• Generate base project structure

Version: 2.0.0

---

2. content-layout-expert

Convert article drafts between formats with image/table extraction.

Description: Converts DOCX/PDF to Markdown with automatic image and table extraction, handles corrupted files via mammoth fallback, and converts Markdown to HTML for WeChat/Xiumi.

Use Cases:

• Convert article drafts (DOCX/PDF) to Markdown
• Extract images to images/ directory
• Extract tables to tables/ directory (XLSX with styling)
• Convert Markdown to HTML for WeChat Official Account
• Handle corrupted DOCX files

Features:

• Auto-heading conversion (first line to #)
• Position markers: (image-N), (table-N)
• Styled XLSX table export
• Preserves text formatting (bold, italic)

Version: 1.0.0

---

3. md-to-pdf

Convert Markdown files to PDF with Chinese font support using WeasyPrint.

Description: Converts Markdown documents to professionally formatted PDFs with full Chinese character support, syntax highlighting, and clean styling.

Use Cases:

• Convert meeting notes to PDF
• Generate documentation PDFs
• Create reports with Chinese text

Features:

• Automatic Chinese font detection (Linux/macOS/Windows)
• Syntax highlighting for code blocks (GitHub Dark theme)
• Table of contents support
• Styled tables, headers, lists
• Page numbers

Version: 1.0.0

---

4. image-stitch

Stitch multiple images together vertically or horizontally.

Description: Combines multiple images into a single composite image with configurable alignment, spacing, and background color.

Use Cases:

• Combine screenshots into one image
• Create photo collages
• Merge document scans
• Stack related images together

Features:

• Vertical or horizontal stitching
• Configurable spacing between images
• Multiple alignment options (left/center/right or top/center/bottom)
• Custom background color support
• Automatic RGB conversion for all image formats

Version: 1.0.0

---

5. tailwind-layout-system

Add complete page layouts and blog system to an existing Astro project.

Description: Enhances an Astro project created by astro-project-init with full page layouts, blog listing with category filtering, and Content Collections configuration.

Use Cases:

• Add complete Home page (Hero, About, Blog, Contact sections)
• Create blog listing page with category filtering
• Add blog post detail pages
• Set up Content Collections

Features:

• Overwrites minimal Home page with complete version
• Blog listing with category filters
• Sample blog posts for testing
• Uses "Late Night Flight" dark theme

Version: 2.0.0

---

Quick Reference

Skill	Primary Function	Input	Output	Documentation
astro-project-init	Initialize Astro project	project_name, target_dir	Complete Astro project structure	SKILL.md
content-layout-expert	Convert document formats	.docx/.pdf/.md file	.md/.html + images/ + tables/	SKILL.md
md-to-pdf	Markdown to PDF	.md file	.pdf file	SKILL.md
image-stitch	Stitch images together	Multiple image files	Single composite image	SKILL.md
tailwind-layout-system	Add layouts to Astro project	project_dir	Enhanced pages + blog system	SKILL.md

---

Skill Relationships

Some skills work together in a workflow:

astro-project-init (foundation)
         ↓
tailwind-layout-system (enhance with layouts)

---

Usage Patterns

Direct Script Execution

# Astro project initialization
bash ~/projects/huang/.claude/skills/astro-project-init/scripts/init.sh my-site .

# Content conversion
source ~/venv/hepmad/bin/activate && python ~/projects/huang/.claude/skills/content-layout-expert/scripts/convert_to_markdown.py input.docx

# Markdown to PDF
source ~/venv/hepmad/bin/activate && python ~/projects/huang/.claude/skills/md-to-pdf/scripts/convert.py input.md

# Image stitching
source ~/venv/hepmad/bin/activate && python ~/projects/huang/.claude/skills/image-stitch/scripts/stitch.py output.jpg img1.jpg img2.jpg

# Add layouts to existing project
bash ~/projects/huang/.claude/skills/tailwind-layout-system/scripts/setup.sh .

Via AI Assistant

Simply describe the task:

• "Initialize a new Astro project"
• "Convert this DOCX to Markdown and extract images"
• "Convert this Markdown file to PDF"
• "Stitch these images together vertically"
• "Add complete page layouts to my Astro project"

---

Directory Structure

.claude/skills/
├── README.md
├── astro-project-init/
│   ├── SKILL.md
│   ├── scripts/
│   │   └── init.sh
│   └── references/
│       └── tailwind-quick-ref.md
├── content-layout-expert/
│   ├── SKILL.md
│   ├── scripts/
│   │   ├── convert_to_markdown.py
│   │   └── convert_to_html.py
│   └── references/
│       └── requirements.txt
├── image-stitch/
│   ├── SKILL.md
│   ├── scripts/
│   │   └── stitch.py
│   └── references/
├── md-to-pdf/
│   ├── SKILL.md
│   ├── scripts/
│   │   └── convert.py
│   └── references/
│       ├── weasyprint-guide.md
│       ├── markdown-extensions.md
│       └── pygments-guide.md
└── tailwind-layout-system/
    ├── SKILL.md
    ├── scripts/
    │   └── setup.sh
    └── references/
        └── astro-content-collections.md

---

Contributing

When adding new skills to this directory:

1. Create a new directory following kebab-case naming
2. Include a SKILL.md file with YAML front matter
3. Add executable scripts in a scripts/ subdirectory
4. Include reference documentation in a references/ subdirectory
5. Update this README with the new skill information

---

License

These skills are part of the Huang project and are intended for personal and educational use.