baoyu-compress-image_skill

Overview

This skill compresses and converts images to WebP (default) or PNG/JPEG using the best available toolchain. It selects the optimal encoder automatically and preserves or replaces originals based on your options. Designed for batch and single-file workflows, it reports size and percentage reductions.

How this skill works

The CLI examines available tools (system sips, cwebp, ImageMagick, or a bundled Sharp fallback) and picks the most effective encoder. It accepts files or directories, can recurse subdirectories, and outputs converted files with new extensions or to a specified output path. Quality, format, and keep-original behavior are configurable and it can emit JSON summaries for automation.

When to use it

• Reduce image file size for web delivery or storage.
• Convert PNG/JPEG assets to WebP for improved performance.
• Optimize large image directories before deployment or publishing.
• Create smaller thumbnails or preview images while keeping originals optional.
• Automate image optimization in build scripts or CI pipelines.

Best practices

• Test quality settings on representative images before batch processing to find the right balance.
• Use recursive mode only on vetted directories to avoid unintended conversions.
• Enable --keep to preserve originals during initial runs; remove originals once satisfied.
• Use --json for programmatic workflows and CI integration.
• Set format explicitly when you must keep PNG or JPEG rather than WebP.

Example use cases

• Compress a single image in place to WebP with default quality (80).
• Batch-convert an images/ folder recursively at quality 75 for web deployment.
• Produce PNG outputs while keeping originals: convert image.png -f png --keep.
• Run the CLI from build scripts and consume its JSON output to report total bytes saved.
• Optimize a media directory before upload to reduce storage and CDN costs.