skills.rest
ludotype
ludotypeCOMMUNITY
Explore all skills by this author →
GitHub

baoyu-cover-image

Generates article cover images using five configurable dimensions and multiple raster backends.

1|Updated May 16, 2026
Search Tags:#prompt engineering#image generation#visual design#color palette#cover image#aspect ratio#article header
One-click install
npx skills add https://github.com/ludotype/game-adventurertown --skill baoyu-cover-image-ludotype
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: baoyu-cover-image
Source: https://github.com/ludotype/game-adventurertown/tree/main/.agent/skills/baoyu-cover-image
Command: npx skills add https://github.com/ludotype/game-adventurertown --skill baoyu-cover-image-ludotype

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Creating consistent, on-brand cover images for articles requires design decisions about composition, color, style, and typography that most writers lack time or skill to make. This Skill turns article content into a structured image-generation prompt and produces a finished raster cover image. ## Core Features & Use Cases - Five-Dimension Customization: Choose type (hero, conceptual, typography, metaphor, scene, minimal), palette (11 options), rendering style (7 options), text density, mood, and font, with content-signal-based auto-selection for any omitted dimension. - Multi-Backend Generation: Resolves an image backend automatically across Codex imagegen, codex-imagegen CLI wrapper, runtime-native tools, or baoyu-imagine, with reference image support via --ref. - Persistent Preferences: Stores watermark, default aspect ratio, output directory, quick mode, and backend pinning in an EXTEND.md file at project or user level. - Use Case: A blogger finishes a technical article and asks for a cover; the Skill analyzes the content, recommends a cool palette with flat-vector rendering, confirms the choices, writes a reproducible prompt file, and outputs a 16:9 cover.png. ## Quick Start Ask the agent to generate a cover image for your article, optionally specifying dimensions like "generate a cover image for this post with the blueprint style in 16:9".

Frequently Asked Questions about baoyu-cover-image

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

FAQPage Schema
How do I generate a cover image for an article?▼

Invoke the skill with a request like "generate cover image" and provide your article content or file path. The skill analyzes the content, recommends type, palette, and rendering dimensions, asks for confirmation, then writes a prompt file and generates cover.png through an available image backend.

What image generation backends does this skill support?▼

It resolves backends in priority order: a user-named override, a saved preferred_image_backend preference, then auto-selection among Codex imagegen, the codex-imagegen CLI wrapper, other runtime-native tools, or baoyu-imagine. If none are available it asks the user how to proceed.

Can I use reference images to match an existing visual style?▼

Yes, pass reference files with --ref or paste them. Files are copied to a refs/ directory and passed to ref-capable backends, while the prompt embeds detailed MUST-level instructions describing brand elements, colors, and layout extracted from the references.

How do I skip the confirmation step for faster generation?▼

Use the --quick flag, say "直接生成" or equivalent wording, or set quick_mode: true in EXTEND.md. The skill then auto-selects all six dimensions from content signals and only asks about aspect ratio unless --aspect is given.

What aspect ratios and output formats are supported?▼

Supported ratios are 16:9 (default), 2.35:1, 4:3, 3:2, 1:1, and 3:4. Output is a raster cover.png placed in a directory determined by the default_output_dir preference: same-dir, imgs-subdir, or an independent cover-image folder.

Why does the skill refuse to fix text in a generated image?▼

The skill prohibits programmatic bitmap patching with tools like ImageMagick or Pillow because overlays degrade quality. If title text is wrong, it regenerates from a corrected prompt file, preserving the flawed candidate for comparison.

Related Skills

Explore More Design & Creative Skills→
XY121718XY121718
101

baoyu-cover-image

Generate customizable article cover images with configurable dimensions and style presets.

Community
Advanced
brainbrewlabsbrainbrewlabs
7

publishing

Format and distribute finalized content to publishing platforms with scheduling and verification.

Official
Intermediate
valtterimelkkovaltterimelkko
1

socialmedia-optimization

Standardize Open Graph and Twitter Card metadata for consistent social media link previews.

Community
Intermediate
AnglesiteAnglesite
2

design-import

Import design tokens and page layouts from Canva or Figma into an Astro site.

Official
Advanced