baoyu-compress-image
🎯Skillfrom jimliu/baoyu-skills
|A utility skill that compresses images to WebP, PNG, or JPEG with automatic tool selection (sips/cwebp/ImageMagick/Sharp), supporting batch processing, configurable quality, and customizable defaults via EXTEND.md.
Overview
Baoyu Compress Image is a utility skill from the jimliu/baoyu-skills collection that compresses images to WebP (default) or PNG format with automatic tool selection. It detects and uses the best available compression tool on the system from sips, cwebp, ImageMagick, or Sharp, ensuring it works across different environments without manual configuration. The skill activates when users request image compression, optimization, WebP conversion, or file size reduction.
Key Features
- Automatic Tool Selection: Detects and uses the best available compression tool in priority order: sips, cwebp, ImageMagick, then Sharp, adapting to whatever is installed on the system
- Multiple Output Formats: Supports WebP (default), PNG, and JPEG output formats with configurable quality levels from 0-100
- Batch Processing: Processes entire directories with optional recursive subdirectory support using the --recursive flag
- Customizable via EXTEND.md: Supports project-level and user-level configuration files for default format, quality, and keep-original preferences
- JSON Output Mode: Provides machine-readable JSON output for integration with automated workflows and pipelines
Who is this for?
- Web developers optimizing images for faster page loads who want one-command compression directly from Claude Code
- Content creators and designers who need to batch-compress images across directories with consistent quality settings
- Teams with mixed development environments who need image compression that automatically adapts to available system tools
Same repository
jimliu/baoyu-skills(27 items)
Installation
npx vibeindex add jimliu/baoyu-skills --skill baoyu-compress-imagenpx skills add jimliu/baoyu-skills --skill baoyu-compress-image~/.claude/skills/baoyu-compress-image/SKILL.mdSKILL.md
More from this repository10
A Claude Code skill for publishing content to WeChat Official Accounts, supporting article posting via API/browser and image-text posting with up to 9 images, with configurable preferences and a structured 7-step publishing workflow.
AI SDK-based image generation skill supporting OpenAI, Google, and DashScope providers with text-to-image generation, reference images, aspect ratio control, and quality presets through a unified slash command interface.
Converts Markdown files to clean, semantic HTML with support for custom styles, code highlighting, and responsive rendering.
A Claude Code skill that generates article cover images using a 5-dimensional system (Type, Palette, Rendering, Text, Mood) with 9 color palettes and 6 rendering styles for 54 unique combinations.
A Claude Code skill that analyzes articles, identifies optimal illustration positions, and generates images using a Type x Style two-dimensional system with 6 types and multiple visual styles.
A Claude Code skill that generates Xiaohongshu (RedNote) infographic card series from content using a Style x Layout 2D system with 9 visual styles and 6 layout types.
A Claude Code skill that transforms written content into professional slide deck images with 16 visual style presets, smart content scaling, and multi-language support.
Generates professional infographics by combining 20 layout types (bento-grid, funnel, timeline, etc.) with 17 visual styles (craft-handmade, cyberpunk, pixel-art, etc.), with auto-recommended combinations based on content analysis.
Fetches any URL via Chrome CDP with full JavaScript rendering and converts it to clean markdown with metadata. Supports auto-capture and wait mode for login-required or dynamically loaded pages.
A Claude Code skill for creating educational and narrative comics with 5 art styles, 7 tones, 3 presets, and 6 layout options, generating detailed panel layouts and sequential images from source material or text input.