- Updated the model fetching logic to include persistent storage for improved performance and reliability.
- Introduced background model list updates to ensure the UI reflects the latest data.
- Enhanced the handling of parameterless tool calls in the AI provider, injecting default arguments when necessary.
- Refactored the chat adapter to support session rollbacks, improving user experience during message regeneration.
These changes significantly improve the functionality and responsiveness of the AI components.
- Refactored type exports in skill-api, runtime-addon, and types files for better clarity and consistency.
- Updated import paths to utilize centralized type definitions, enhancing maintainability.
- Improved the BrowserMessageActions component to handle text content more effectively when copying messages.
These changes contribute to a cleaner codebase and improved functionality across the project.
- Adjusted formatting in SettingsPage for better readability.
- Streamlined import statements in browser-chat-header and browser-message-actions for consistency.
- Enhanced the MCP bridge panel layout for improved clarity.
- Refactored WebSocket bridge exports for better structure.
These changes enhance code maintainability and readability across the project.
- Updated ChatAdapter to handle new assistant messages after tool calls, improving message management.
- Added connection tab support in SettingsPage, allowing for better organization of settings.
- Introduced connectionContent prop in SettingsPageProps for dynamic content rendering.
- Enhanced message-item component to include onLogin handler for improved user interaction.
- Refactored useChat and useChatConfig hooks for better state management and silent reload capabilities.
- Implemented WebSocket MCP bridge in background script for real-time communication.
This update improves user experience and functionality across chat and settings components.
- Streamlined code formatting in multiple files for better consistency and clarity.
- Consolidated import statements and removed unnecessary line breaks to enhance organization.
- Updated state initialization and prop handling in various components to follow best practices.
- Improved the readability of complex expressions by simplifying their structure.
- Added functionality to extract and apply screenshots from tool results in the ChatAdapter, improving user experience by allowing immediate rendering of screenshots.
- Implemented a new ToolScreenshot component to display screenshots inline, supporting both base64 data and IndexedDB references.
- Updated message item rendering to transform screenshot placeholders into actual images, enhancing the visual feedback for users.
- Introduced collapsible message items for intermediate assistant messages, improving the organization of conversation turns in the message list.
- Enhanced model fetching logic in the Chatbot component to ensure server models are prioritized, improving model selection reliability.
- Updated localization files to include new translation keys for improved user guidance.
- Updated the status and impact of the `organize_tabs` tool in LOGIC_INCONSISTENCIES_BY_PACKAGE.md, noting its removal from the default bundle.
- Resolved naming inconsistencies for the `ungroup_tabs` tool in the tab-groups module.
- Implemented the `refreshSkillMetadata` method in SkillManager to update skill metadata from SKILL.md, including validation and error handling.
- Reformatted object literals in tests for better readability.
- Cleaned up imports in TokenUsageIndicator component.
- Enhanced message-adapter logic to handle missing toolCallId.
- Updated vitest configuration to exclude Puppeteer tests.
- Added type definitions for tab group colors in organize-tabs.
- Refactored color handling in organize-tabs to use new type.
- Improved sanitization functions for better character handling.
- Adjusted test cases for consistency in rawResponses structure.
- Updated ChatAdapter to maintain "streaming" status for tool errors instead of setting it to "error".
- Added logic to extract business-level failure messages from tool results.
- Introduced metrics tracking in the chat context, including updates on token usage.
- Enhanced tests to verify correct handling of tool errors and metrics updates.
- Updated UI components to display metrics and error states appropriately.
