image-loading_skill

Overview

This skill generates a production-ready image loading pipeline for Swift apps, including memory and disk caching, request deduplication, optional image processing, and a drop-in CachedAsyncImage SwiftUI view. It targets modern Swift and platform APIs to replace AsyncImage with a cached, performant alternative. The output includes configurable cache sizes, LRU disk behavior, and optional prefetching and processing components.

How this skill works

I create a small set of Swift components that work together: a memory cache (NSCache) and an LRU disk cache, an actor-based downloader that deduplicates concurrent requests and supports cancellation, an image pipeline that orchestrates cache → download → process → store, and a CachedAsyncImage view for SwiftUI. Optional modules include an ImageProcessor for resizing/thumbnails and an ImagePrefetcher to warm caches for lists. The pipeline uses platform abstractions so it can target iOS (UIImage), macOS (NSImage), or both via a typealias.

When to use it

• You need image caching to avoid repeated network downloads
• You want a drop-in replacement for AsyncImage with caching and cancellation
• You need request deduplication to prevent parallel downloads for the same URL
• You want thumbnail generation or resizing to reduce memory use
• You need prefetching for collection views or lists

Best practices

• Choose medium cache sizes for general apps; increase only if profiling shows need
• Enable image processing (downscale) for large source images to save memory and disk
• Use the prefetcher for long lists but limit concurrency and buffer size
• Respect cached expirations and provide a clear cache invalidation strategy
• Test deduplication and cancellation paths with mocked downloaders in unit tests

Example use cases

• Replace AsyncImage in a profile list to add caching and faster scrolling
• Add thumbnail generation for a photo gallery to reduce memory and disk usage
• Prefetch images for an infinite-scrolling feed to improve perceived performance
• Use CachedAsyncImage in a chat app for avatars with quick cache hits and low bandwidth
• Run unit tests that assert cache hits avoid network downloads and deduplicate concurrent requests