skills.rest
cat-xierluo
cat-xierluoCOMMUNITY
Explore all skills by this author →
GitHub

handdrawn-article-illustrator

Generate hand-drawn editorial illustrations for WeChat articles using theme-based prompts and built-in image generation.

636|91|Updated Jan 7, 2026
Search Tags:#image generation#article#theme#illustration#prompt#hand-drawn#weChat
One-click install
npx skills add https://github.com/cat-xierluo/legal-skills --skill handdrawn-article-illustrator
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: handdrawn-article-illustrator
Source: https://github.com/cat-xierluo/legal-skills/tree/main/skills/handdrawn-article-illustrator
Command: npx skills add https://github.com/cat-xierluo/legal-skills --skill handdrawn-article-illustrator

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pillow, requests, pngquant, imagemagick, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

Creating a consistent set of illustrations for long-form WeChat articles, blogs, or newsletters is time-consuming, and ad-hoc image prompts often produce inconsistent styles, readable text artifacts, or off-brand colors. This Skill turns an article outline into a unified hand-drawn illustration system with controlled composition, palette, and abstraction level.

Core Features & Use Cases

  • Article-aware illustration planning: Reads article structure (headings, transitions, case sections) and plans a cover image plus high-density inline illustrations, each carrying one concept with a written Image Brief and English visual prompt.
  • Theme pack system: Colors (accent, paper, ink) are decoupled into JSON theme files under themes/ (blue-gray, ink-black, terracotta), switchable via CLI flag, outline field, or config, with support for custom firm/brand themes.
  • Controlled generation pipeline: Scripts compile outlines into prompts and a Codex/ImageGen generation queue, enforce composition-family diversity and layout audits, and compress final PNGs below 500KB for delivery.
  • Use Case: Given a 5,000-word legal analysis article, produce 1 cover image plus 12-16 inline illustrations in a consistent hand-drawn object-card style, each mapped to a section's core argument.

Quick Start

Ask the AI to read your article and generate a hand-drawn illustration plan with prompts, then run the batch script with your outline JSON to produce the built-in image generation queue.

Frequently Asked Questions about handdrawn-article-illustrator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate illustrations for a WeChat article?▼

Prepare an outline JSON with your article title and an illustrations array listing each image's position, metaphor, and concept. Run batch_article_images.py with --mode codex to produce a generation queue, then the agent calls the built-in image generation tool for each item to create final PNGs.

How do I change the accent color or create a custom theme?▼

Colors live in JSON theme files under themes/. Create a new file like themes/my-firm.json defining accent_color, paper_background, and ink_color, then activate it with the --theme CLI flag, the outline style.theme field, or config/style.json's active_theme setting.

Does this skill call external image generation APIs by default?▼

No. The default route uses the host environment's built-in image generation via a codex queue. External APIs like SiliconFlow, Z-Image, or Qwen are explicit experimental modes that require the --allow-api flag and a configured SILICONFLOW_AK key.

What image sizes and formats does it produce?▼

Covers are generated wide and exported at 2400x1024 with center-safe composition, inline images are 16:9, and square cards are 1024x1024. Final PNGs are compressed in place by compress_images.py to under 500KB, targeting 200KB.

Why do generated images sometimes contain readable text or UI icons?▼

Image models tend to render article titles, product names, or terms like roles and permissions as literal text or icons. The skill prevents this by rewriting prompts into visual-only concepts and banning checkmarks, avatars, arrows, and readable words in negative prompts.

What are the limitations of the local rendering mode?▼

The --mode local renderer only produces diagnostic composition sketches, not final deliverables. It should run only when a user explicitly requests placeholder previews, and its output must be labeled as a draft rather than a finished illustration.

Related Skills

Explore More Content & Communication Skills→
minicooheiminicoohei
345

article-writer

Generate structured Markdown articles from themes with outlines and illustration markers.

Community
Advanced
alchaincyfalchaincyf
1.5k

huashu-image-upload

Generates article illustrations and uploads them to ImgBB for permanent Markdown links.

Community
Intermediate
ByungJu-LimByungJu-Lim

theme-concept

Translate concepts into text-free Kurzgesagt-inspired visuals with nine mood palettes.

Community
Advanced
darblexdarblex

Image Generation

Generate AI images from text prompts with style and provider control.

Community
Advanced