• chore: bump version to 0.5.0
• feat: update README and onboarding guide for multi-provider support
• feat: add experimental feature tooltip and enhance provider selection UI
• refactor: enhance language configuration in provider settings
• fix: update content security policy to include Google Generative Language API
• fix: remove debug flag from side panel URL in background script
• feat: replace react-feather with lucide-react for icon components
• refactor: update MainPanel and SettingsPanel for improved settings management
• feat: consolidate assistant response handling in GeminiClient
• feat: enhance conversation handling in GeminiClient
• refactor: optimize models API requests to prevent duplicates
• feat: optimize API calls by merging validation and model fetching
• fix: correct event type extraction in OpenAIClient
• refactor: group outputTranscription with modelTurn events
• chore: remove obsolete commit message file
• feat: add support for Gemini input and output transcription messages
• fix: ensure proper session cleanup when WebSocket connection closes
• feat: add log grouping support for Gemini realtime events
• feat: add modelsRequired and modelsLoading translations for all 30 languages
• feat(settings): enhance API key validation and model fetching logic
• feat(log): enhance realtime event handling and logging structure
• feat(gemini): add speech configuration to GeminiClient for voice support
• feat(settings): update settings management to use FilteredModel and enhance language handling
• feat(translations): update language options in GeminiProviderConfig
• feat(translations): add provider and API key translations across multiple languages - Update translation files to include new keys for provider, provider type, and API key in various languages.
• fix(translations): update real-time model availability messages across multiple languages - Simplify translations for real-time model availability by removing specific references to GPT-4o in favor of a more general term.
• feat(gemini): enhance API key validation and model fetching - Implement pagination support for fetching models from Gemini API, improve validation logic to check for realtime model availability, and update interfaces to reflect changes in model structure.
• feat(settings): enhance model selection in ProviderSpecificSettings
• feat(settings): restructure settings management to support common and provider-specific configurations - Introduce CommonSettings, OpenAISettings, and GeminiSettings interfaces for better organization. Update SettingsContext and components to utilize new structure, ensuring backward compatibility with legacy settings. Enhance SettingsPanel and ProviderSpecificSettings to reflect changes in state management and API key validation.
• feat(clients): implement ClientFactory and client adapters for OpenAI and Gemini - Introduce ClientFactory for creating AI client instances, and add OpenAIClient and GeminiClient implementations to handle respective API interactions. Define IClient interface for unified client management.
• refactor: merge service provider & api key sections and refactor validation logic - Merge Service Provider and API Key sections in SettingsPanel - Refactor ApiKeyValidator functionality into client classes - Add static validation methods to OpenAIClient and GeminiClient - Update BrowserSettingsService and ElectronSettingsService to use client methods - Enhance provider configuration system with 30 Gemini voices - Remove ApiKeyValidator.ts utility class - Improve UI layout with simplified section titles and labels
• chore(deps): update dependencies and bump version to 0.4.6 - Added @google/genai and updated package-lock.json with new dependencies including google-auth-libuuuuuuuuuuuuu
• feat(pulseaudio): enhance audio port connection handling
• refactor(webpack): remove commented alias description in webpack configuration - Clean up the webpack.config.js file by removing an unnecessary comment for better readability.