What problem does it solve? SwiftUI code often escapes auto-localization in subtle ways: String variables passed to Text, verbatim: misuse, sentences built by interpolation that break pluralization, locale-unaware number and date formatting, and RTL-unsafe layout. This Skill systematically detects these defects in a macOS SwiftUI project and writes structured findings to disk. ## Core Features & Use Cases - Ten-rule defect index (loc-01 to loc-10): Detects hardcoded strings, verbatim misuse, missing String Catalogs, missing translator comments, interpolation-built sentences, locale-unaware formatting, and RTL-unsafe symbols and offsets. - Hybrid lint engine: Combines tier-1 grep tells with a tier-2 ast-grep structural rule that catches non-literal String arguments passed to Text, emitting unified JSON and SARIF output. - Evidence-grounded verification: Cross-checks uncertain findings against a corpus of 1,857 shipping macOS apps via swiftui-ctx and Apple documentation via Sosumi before reporting. - Use Case: Run it on an in-progress macOS SwiftUI app before release to find every string that will never reach a translator, then review flag-only findings organized by context folders like hardcoded-strings, plurals-and-grammar, and rtl-layout. ## Quick Start Audit the SwiftUI sources in this project for localization and internationalization problems and write the findings to swiftui-audits/localization.
