natural-language

The natural-language skill provides on-device text analysis and language translation capabilities for iOS, macOS, and visionOS applications using Apple's NaturalLanguage and Translation frameworks. It enables developers to add intelligent text processing features without sending data to external servers, ensuring user privacy while performing operations like tokenization, language detection, sentiment analysis, named entity recognition, and in-app translation. This skill targets Swift 6.3 and iOS 26+, offering a modern approach to natural language processing in Apple ecosystem apps.

The skill encompasses two primary frameworks: NaturalLanguage for text analysis operations and Translation for language conversion. NaturalLanguage provides tokenization capabilities (breaking text into words, sentences, or paragraphs), part-of-speech tagging, named entity recognition to identify people, places, and organizations, sentiment analysis to determine emotional tone, language identification supporting multiple hypotheses with confidence scores, and text embeddings for semantic similarity comparisons. The Translation framework enables in-app language translation with both UI-based and programmatic approaches, supporting batch translation and language availability checks.

This skill is ideal for developers building content analysis tools, multilingual applications, text editors with intelligent features, social media apps requiring sentiment analysis, search systems using semantic understanding, or any application that needs to process and understand natural language text. It clearly defines scope boundaries, handling text analysis after acquisition while delegating OCR to vision-framework, speech-to-text to speech-recognition, and generative AI tasks to apple-on-device-ai. The skill emphasizes thread-safety considerations and provides practical examples for common use cases.