PPQS
WILD CORPUS · github_awesome

PQS 52 (C) - prompt from raw.githubusercontent.com

Source: raw.githubusercontent.com · Scraped 2026-05-04 · Scored 2026-05-04

Score

C
52 / 80
gemma4:latest · local · pqs-v2.0 · canonical
Clarity9 / 10
Specificity9 / 10
Context8 / 10
Constraints7 / 10
Output format5 / 10
Role definition10 / 10
Examples1 / 10
CoT structure3 / 10

The prompt

Act as an App Localization Expert. You are tasked with setting up a user-preference-based localization architecture in an application independent of the phone's system language.

Your task includes:
1. **LanguageManager Class**: Create a `LanguageManager` class using the `ObservableObject` protocol. Store the user's selected language in `UserDefaults`, with the default language set to 'en' (English). Display a selection screen on the first launch.
2. **Global Locale Override**: Wrap the entire `ContentView` structure in your SwiftUI app with `.environment(\.locale, .init(identifier: languageManager.selectedLanguage))` to trigger translations based on the selected language in `LanguageManager`.
3. **Onboarding Language Selection**: If no language has been selected previously, show a stylish 'Language Selection' screen with English and Turkish options on app launch. Save the selection immediately and transition to the main screen.
4. **AI (LLM) Integration**: Add the user's selected language as a parameter in AI requests (API calls). Update the system prompt to: 'User's preferred language: ${selected_language}. Respond in this language.'
5. **String Catalogs**: Integrate `.stringxcatalog` into your project and add all existing hardcoded strings in English (base) and Turkish.
6. **Dynamic Update**: Ensure that changing the language in settings updates the UI without restarting the app.
7. **User Language Change**: Allow users to change the app's language dynamically at any time.

Rules:
- Ensure seamless user experience during language selection and updates.
- Test functionality for both English and Turkish languages.

This prompt was scraped from a public source. The score reflects the input as written, not the quality of any output it produced. The AI input quality problem is the gap between what people type and what the model can act on.