Apple's NaturalLanguage and Translation frameworks give you on-device text analysis and language translation without sending data to the cloud. You get tokenization, part-of-speech tagging, named entity recognition, sentiment scoring, and text embeddings for semantic similarity. The Translation framework adds both a system UI overlay and programmatic translation with batch support. NLTagger and NLTokenizer aren't thread-safe, so keep instances on one queue. Sentiment scores range from negative one to positive one, and embeddings work at both word and sentence level. Useful when you need language features in iOS or macOS apps without the latency or privacy concerns of API calls.
npx skills add https://github.com/dpearson2699/swift-ios-skills --skill natural-language