About this skill

Problem to solve

When PNG screenshots, assets, or batched images are too large, manual compression and repeated setting tweaks are tedious. image-compression focuses on compressing PNG files: provide a file path or directory and get a measurable size reduction.

How it works

Lossless mode: uses optimize=True and compress_level=9, typically reducing size by 1-30%. It suits screenshots, UI captures, and text-heavy images where sharpness matters.
Lossy mode: quantizes to 256 colors with Image.Quantize.MEDIANCUT, then saves with optimize=True, typically reducing size by 50-80%. It suits photos, gradients, or images where smaller file size is the priority.
Batch processing: pass a directory to process PNGs in a folder. compress_png.py supports arguments such as --output, --overwrite, and --lossy, making it easier to write results to a target path or replace originals.
Result reporting: after compression, it reports original size, compressed size, and savings percentage for quick validation.

Scope and cautions

This skill covers PNG compression only; it does not resize, reformat, or alter image content. If the user does not specify a mode, ask whether lossless or lossy is preferred. For color-critical images or images with many color layers, prefer lossless mode. Lossy mode reduces the number of colors, so visual results depend on the image content.

Screenshots

Use Cases

Compress an oversized PNG screenshot before sending it to a collaborator, choosing lossless or lossy based on the need.
Batch-compress PNG assets in a folder, write results to an output directory, and report savings.
Keep text and UI edges sharp in a PNG screenshot by using lossless compression and checking the reduction.
Reduce photo-like PNGs with 256-color lossy compression when visual changes are acceptable.

Best For

Frontend or QA engineers who need to compress UI screenshots for bug reports or docs.
Design collaborators who need to shrink multiple PNG assets before packaging or sharing.
Data labeling or asset-curation staff who need batch PNG compression with output tracking.
Indie developers who need to reduce PNG screenshot or demo image weight quickly.