translation-extract

What problem does it solve? Hardcoded strings scattered across a codebase make internationalization painful and error-prone. This Skill scans source files to find untranslated user-facing text, proposes consistent translation keys, and optionally rewrites the code to use your i18n framework's translation functions. ## Core Features & Use Cases - Framework Detection: Automatically identifies your i18n setup (react-intl, i18next, vue-i18n, Angular, svelte-i18n, Flutter, Rails, Django, Laravel) from project dependency files. - Confidence-Rated Extraction: Classifies findings as high or medium confidence while excluding logs, test files, CSS classes, URLs, and other non-translatable strings. - Key Generation and Rewrite: Proposes snake_case keys with contextual prefixes, and with the --write flag adds keys to the source language file and replaces hardcoded strings with translation function calls including imports. - Use Case: You are preparing a React app for launch in new markets. Run the scan on src/ to get a report of 12 hardcoded strings with proposed keys, then apply --write to update en.json and the components automatically. ## Quick Start Scan the src directory for hardcoded user-facing strings and show me a report with proposed translation keys.