Smart Image Compressor

Problem

Images for uploads, chat attachments, and blog posts are often too large. A 4000x3000 photo or high-resolution screenshot can slow down APIs and increase bandwidth costs, while manually tuning format and quality can easily make text screenshots blurry.

How It Works

The skill uses Pillow to compress images. Instead of applying a fixed quality=50, it selects a strategy based on target compression ratio and format behavior:
- JPEG quality tuning: suits photos; searches the quality range and can optionally remove EXIF.
- PNG optimization: preserves transparency where possible, suggests JPEG conversion for opaque PNGs, and optimizes indexed palettes.
- WebP conversion: can reduce size further at comparable quality, but compatibility with older platforms such as WeChat should be considered.
- Proportional resizing: oversized images can be scaled first using --max-width before compression.

It can generate a compression report and, if needed, read before and after images for visual comparison. By default it writes original-name_compressed.ext, preserving the source file.

Boundaries

It fits common cases where size must drop without obvious quality loss, especially photos, screenshots, UI assets, and WebP output. If the source is already under 100KB, gains may be limited. Transparent PNG to JPEG requires a white background fill, and users should be told explicitly. Very low quality settings can cause visible artifacts.