baoyu-cover-image_skill

This skill generates article cover images with 5-dimension customization (type, palette, rendering, text, mood) across aspect ratios to enhance engagement.
  • TypeScript

5.8k

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill jimliu/baoyu-skills --skill baoyu-cover-image

  • SKILL.md9.1 KB

Overview

This skill generates elegant article cover images with five-dimensional customization: type, palette, rendering, text, and mood. It combines nine color palettes and six rendering styles and supports cinematic, widescreen, and square aspect ratios. Use natural commands like "generate cover image", "create article cover", or "make cover" to invoke it. Outputs include a prompt file, saved references, and a final cover image in a structured output directory.

How this skill works

The tool analyzes the article content and any supplied reference images to auto-select or confirm six cover dimensions: type, palette, rendering, text level, mood, and font. It builds a detailed generation prompt (saved to prompts/cover.md), processes references into either direct attachments or extracted style traits, and calls an image-generation backend to produce the cover. Workflows include a quick mode to skip confirmations, automatic retries on failure, and safe backup of existing cover files before regeneration.

When to use it

  • You need a polished cover image for a blog post, newsletter, or report.
  • You want consistent multi-article visual language using presets or EXTEND.md preferences.
  • You have reference images or a specific visual direction to incorporate.
  • You need different aspect ratios for platforms (cinematic 2.35:1, 16:9, 1:1).
  • You prefer automated sensible defaults but want manual overrides for dimensions.

Best practices

  • Provide the exact article title in the source; the title is used verbatim and never invented.
  • Include clear reference images when you want direct style or composition reuse; the skill treats them as MUST elements when referenced.
  • Use --quick for fast generation when defaults are acceptable; otherwise confirm the six dimensions to tailor results.
  • Set EXTEND.md in project or user config on first run to avoid blocking setup prompts and to store preferred defaults.
  • Backup or rename existing cover.png before regenerating if you want to preserve prior versions.

Example use cases

  • Generate a hero-style cinematic cover for a long-form feature article with a warm palette and painterly rendering.
  • Create a square social-media cover with minimal typography and a pastel palette for an interview post.
  • Batch-create covers for a series using a style preset (shorthand) to maintain consistent palette and rendering.
  • Produce a conceptual cover guided by two reference images: one for composition, one for color palette extraction.
  • Quickly make a text-rich cover for a tutorial by specifying --text text-rich and --aspect 16:9.

FAQ

Use commands like "generate cover image", "create article cover", or "make cover"; CLI aliases and flags also invoke specific behaviors.

How do I force no title?

Use --no-title or --text none to produce a purely visual cover with no overlayed text.

Where are outputs saved?

Outputs follow your default_output_dir preference: same-dir, imgs-subdir, or independent (cover-image/{topic-slug}/) and include source, refs, prompts/cover.md, and cover.png.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational