* refactor: use code font family in HtmlArtifactsCard
* fix: pass onSave to HtmlArtifactsPopup
* feat: add a save button
* fix: avoid extra blank lines
* feat: make split view resizable
* refactor: improve streaming check, simplify Markdown component
* refactor: improve button style and icons
* test: update snapshots, add tests
* refactor: move font family to TerminalPreview
* test: update
* refactor: add explicit type for Node
* refactor: remove min-height
* fix: type
* refactor: improve scrollbar and splitter style
* fix: prevent incorrect @ symbol deletion in QuickPanel
- Track trigger source (input vs button) and @ position
- Only delete @ when triggered by input with model selection
- Button-triggered panels never delete text content
- Validate @ still exists at recorded position before deletion
* feat: delete search text along with @ symbol
- Pass searchText from QuickPanel to onClose callback
- Delete both @ and search text (e.g., @cla) when model selected
- Validate text matches before deletion for safety
- Fallback to deleting only @ if text doesn't match
* refactor: clarify ESC vs Backspace behavior in QuickPanel
- ESC: Cancel operation, delete @ + searchText when models selected
- Backspace: Natural editing, @ already deleted by browser, no extra action
- Clear separation of intent improves predictability and UX
* fix(export): robustly export reasoning and handle errors
* fix(export): normalize <br> to newline before notion parsing
* feat(i18n): add notion truncation and unify export warn keys
* refactor(export): add typing, state guards, and error logging
* fix(export): preserve existing <br> in reasoning when convert to html
* feat(export): add DOMPurify sanitization for reasoning content
* chore(deps): remove unused @types/dompurify dev dep
* chore(deps): remove dompurify dependency
Remove dompurify from package.json and yarn. The changes
delete dompurify entries and simplify the lockfile resolution so the
project no longer declares dompurify as a direct dependency.
This cleans up unused dependency declarations and prevents installing
dompurify when it is not required.
* refactor(ManageModelsPopup): remove margin of Empty
* chore: use destroyOnHidden rather than deprecated destroyTooltipOnHide
* refactor: center Empty
* fix: Update KnowledgeBase form and service to handle preprocess provider correctly
- Enhanced useKnowledgeBaseForm hook to set preprocessProvider with the correct providerId type.
- Modified getKnowledgeBaseParams function to retrieve preprocess provider from the store instead of the base, ensuring accurate provider assignment.
* fix: Remove unused providerId from preprocessProvider in useKnowledgeBaseForm hook
- Cleaned up the useKnowledgeBaseForm hook by removing the unused providerId property from the preprocessProvider object, ensuring a more accurate representation of the data structure.
* format code
* feat: Sync preprocess provider updates across knowledge bases
- Added a new action to synchronize updates to the preprocess provider across all knowledge bases that reference it.
- Updated the usePreprocessProvider hook to dispatch the new sync action after updating the provider.
- Modified getKnowledgeBaseParams to ensure the correct preprocess provider is assigned when retrieving knowledge base parameters.
* fix: Update sync logic for preprocess provider updates
- Modified the syncPreprocessProvider action to merge updates directly into the existing provider object instead of replacing it.
- Adjusted the usePreprocessProvider hook to only dispatch the sync action when specific fields (apiHost, apiKey, model) are updated, improving efficiency.
* fix(QuickPanel): auto-close panel when no commands match (#7824)
Fixes the issue where QuickPanel remains visible when user types
invalid slash commands. Now the panel intelligently closes after
300ms when no matching commands are found.
- Add smart delayed closing mechanism for unmatched searches
- Optimize memory management with proper timer cleanup
- Preserve existing trigger behavior for / and @ symbols
* feat(inputbar): intelligent @ symbol handling on model selection panel close
- Add smart @ character deletion when user selects models and closes panel with ESC/Backspace
- Preserve @ character when user closes panel without selecting any models
- Implement action tracking using useRef to detect user model interactions
- Support both ESC key and Backspace key for consistent behavior
- Use React setState instead of DOM manipulation for proper state management
Resolves user experience issue where @ symbol always remained after closing model selection panel
* perf(QuickPanel): optimize timer management and fix React anti-patterns
- Move side effects from useMemo to useEffect for proper React lifecycle
- Add automatic timer cleanup on component unmount and dependency changes
- Remove unnecessary timer creation/destruction on each search input
- Improve memory management and prevent potential memory leaks
- Maintain existing smart auto-close functionality with better performance
Fixes React anti-pattern where side effects were executed in useMemo,
which should be a pure function. This improves performance especially
when users type quickly in the search input.
* refactor(QuickPanel): remove redundant timer cleanup useEffect
Remove duplicate timer cleanup logic as the existing useEffect at line 141-164
already handles component unmount cleanup properly.
* refactor(QuickPanel): optimize useEffect dependencies and timer cleanup logic
- Replace overly broad `ctx` dependency with precise `[ctx.isVisible, searchText, list.length, ctx.close]`
- Move timer cleanup before visibility check to ensure proper cleanup on panel hide
- Add early return when panel is invisible to prevent unnecessary timer creation
- Improve performance by avoiding redundant effect executions
- Fix edge case where timers might not be cleared when panel becomes invisible
Addresses review feedback about dependency array optimization while maintaining
existing auto-close functionality and improving memory management.
* feat(QuickPanel): implement smart re-opening with dependency optimization
Features:
- Implement smart re-opening during deletion with real-time matching
- Only reopen panel when actual matches exist to avoid unnecessary interactions
- Add intelligent @ symbol handling on model selection panel close
- Optimize search text length limits (≤10 chars) for performance
Fixes:
- Fix useMemo dependency from overly broad [ctx, searchText] to precise [ctx.isVisible, ctx.symbol, ctx.list, searchText]
- Resolve trailing whitespace formatting issues
- Eliminate ESLint exhaustive-deps warnings while maintaining stability
- Prevent unnecessary re-renders when unrelated ctx properties change
Performance improvements ensure optimal QuickPanel responsiveness while maintaining
existing auto-close functionality and improving user experience.
* fix(ci): add eslint-disable comment for exhaustive-deps warning
The useEffect dependency array [ctx.isVisible, searchText, list.length, ctx.close]
is intentionally precise to avoid unnecessary re-renders when unrelated ctx
properties change. Adding ctx object would cause performance degradation.
* refactor(QuickPanel): remove smart re-opening logic during deletion
- Remove 62 lines of complex deletion detection logic from Inputbar component
- Eliminates performance overhead from frequent string matching during typing
- Reduces code complexity and potential edge cases
- Maintains simple and predictable QuickPanel behavior
- Improves maintainability by removing unnecessary "smart" features
The deletion-triggered smart reopening feature added unnecessary complexity
without significant user benefit. Users can simply type / or @ again to
reopen panels when needed.
- Introduced QuickPhraseSettings component for managing quick phrases with add, edit, and delete functionality.
- Added PreprocessSettings component to configure document preprocessing options, including provider selection.
- Updated SettingsPage to include links to the new Quick Phrase and Preprocess settings.
- Removed the deprecated ToolSettings component and its associated routes.
- Enhanced WebSearchSettings with new compression settings and improved UI for managing web search providers.
* feat: add code tools
* feat(CodeToolsService): add CLI executable management and installation check
- Introduced methods to determine the CLI executable name based on the tool.
- Added functionality to check if a package is installed and create the necessary bin directory if it doesn't exist.
- Enhanced the run method to handle installation and execution of CLI tools based on their installation status.
- Updated terminal command handling for different operating systems with improved comments and error messages.
* feat(ipService): implement IP address country detection and npm registry URL selection
- Added a new module for IP address country detection using the ipinfo.io API.
- Implemented functions to check if the user is in China and to return the appropriate npm registry URL based on the user's location.
- Updated AppUpdater and CodeToolsService to utilize the new ipService functions for improved user experience based on geographical location.
- Enhanced error handling and logging for better debugging and user feedback.
* feat: remember cli model
* feat(CodeToolsService): update options for auto-update functionality
- Refactored the options parameter in CodeToolsService to replace checkUpdate and forceUpdate with autoUpdateToLatest.
- Updated logic to handle automatic updates when the CLI tool is already installed.
- Modified related UI components to reflect the new auto-update option.
- Added corresponding translations for the new feature in multiple languages.
* feat(CodeToolsService): enhance CLI tool launch with debugging support
- Added detailed logging for CLI tool launch process, including environment variables and options.
- Implemented a temporary batch file for Windows to facilitate debugging and command execution.
- Improved error handling and cleanup for the temporary batch file after execution.
- Updated terminal command handling to use the new batch file for safer execution.
* refactor(CodeToolsService): simplify command execution output
- Removed display of environment variable settings during command execution in the CLI tool.
- Updated comments for clarity on the command execution process.
* feat(CodePage): add model filtering logic for provider selection
- Introduced a modelPredicate function to filter out embedding, rerank, and text-to-image models from the available providers.
- Updated the ModelSelector component to utilize the new predicate for improved model selection experience.
* refactor(CodeToolsService): improve logging and cleanup for CLI tool execution
- Updated logging to display only the keys of environment variables during CLI tool launch for better clarity.
- Introduced a variable to store the path of the temporary batch file for Windows.
- Enhanced cleanup logic to remove the temporary batch file after execution, improving resource management.
* feat(Router): replace CodePage with CodeToolsPage and add new page for code tools
- Updated Router to import and route to the new CodeToolsPage instead of the old CodePage.
- Introduced CodeToolsPage component, which provides a user interface for selecting CLI tools and models, managing directories, and launching code tools with enhanced functionality.
* refactor(CodeToolsService): improve temporary file management and cleanup
- Removed unused variable for Windows batch file path.
- Added a cleanup task to delete the temporary batch file after 10 seconds to enhance resource management.
- Updated logging to ensure clarity during the execution of CLI tools.
* refactor(CodeToolsService): streamline environment variable handling for CLI tool execution
- Introduced a utility function to remove proxy-related environment variables before launching terminal processes.
- Updated logging to display only the relevant environment variable keys, enhancing clarity during execution.
* refactor(MCPService, CodeToolsService): unify proxy environment variable handling
- Replaced custom proxy removal logic with a shared utility function `removeEnvProxy` to streamline environment variable management across services.
- Updated logging to reflect changes in environment variable handling during CLI tool execution.
- Introduced ApiOptionsSettings component to manage API options for providers.
- Added ApiOptionsSettingsPopup for displaying API options in a modal.
- Updated ProviderSetting to include a button for opening the API options popup for non-system providers.
- Refactored imports and adjusted layout in ProviderSetting for better UI consistency.
- Simplified proxy setting logic by removing unnecessary dispatches for 'system' and 'none' modes.
- Updated useAppInit to set proxy to undefined for 'system' mode and clarified direct mode handling with comments.
- Removed TranslateSettings and TranslateModelSettings components to streamline the translation settings interface.
- Introduced CustomLanguageSettings and TranslatePromptSettings components for better management of custom languages and prompt settings.
- Updated ModelSettings to utilize the new TranslateSettingsPopup for handling translation-related configurations.
- Enhanced the overall structure and readability of the translation settings page.
* refactor(FileManager): streamline file path handling in FilesPage and ImageBlock components
* refactor(file): implement getSafeFilePath utility for consistent file path handling across loaders and preprocessors
* refactor(FileStorage): replace getSafeFilePath with fileStorage.getFilePathById for consistent file path retrieval across services
* refactor(file): unify file path retrieval across loaders and preprocessors for improved consistency
* refactor(Inputbar, MessageEditor): replace getFileExtension with file.ext for improved file type handling
* refactor(FileStorage): simplify getFilePathById method by removing redundant checks for file path retrieval
* fix(FileStorage): update getFilePathById to ensure file.path is consistent with generated filePath
* refactor(FileStorage): simplify getFilePathById method by removing unnecessary file path consistency checks
* fix(FileStorage): update duplicate file check to use file.path for accurate detection
* fix(FileStorage): correct file path usage in uploadFile method for accurate duplicate detection
* fix(loader): update file path retrieval to use file.path for consistency across loaders
- Removed unused imports and components to simplify the codebase.
- Refactored the token count calculation for improved readability.
- Adjusted the layout of the operation bar and input/output containers for better spacing and alignment.
- Enhanced the copy button functionality and visibility within the output area.
- Updated styles for consistency and improved user experience.
* Update models.ts
* Update models.ts
* Update models.ts
* feat: add OpenAI verbosity setting for GPT-5 model
Introduces a new 'verbosity' option for the OpenAI GPT-5 model, allowing users to control the level of detail in model output. Updates settings state, migration logic, UI components, and i18n translations to support this feature.
* fix(models): 修正gpt-5模型判断逻辑以支持包含gpt-5的模型ID
* fix(i18n): 修正繁体中文和希腊语的翻译错误
* fix(models): 优化OpenAI推理模型判断逻辑
* fix(OpenAIResponseAPIClient): 不再为response api添加stream_options
* fix: update OpenAI model check and add verbosity setting
Changed GPT-5 model detection to use includes instead of strict equality. Added default 'verbosity' property to OpenAI settings in migration logic.
* feat(models): 添加 GPT-5 系列模型的图标和配置
添加 GPT-5、GPT-5-chat、GPT-5-mini 和 GPT-5-nano 的图标文件，并在 models.ts 中配置对应的模型 logo
* Merge branch 'main' into fix-gpt5
* Add verbosity setting to OpenAI API client
Introduces a getVerbosity method in BaseApiClient to retrieve verbosity from settings, and passes this value in the OpenAIResponseAPIClient request payload. This enables configurable response verbosity for OpenAI API interactions.
* Upgrade OpenAI package to 5.12.2 and update patch
Replaced the OpenAI dependency from version 5.12.0 to 5.12.2 and updated related patch files and references in package.json and yarn.lock. Also updated a log message in BaseApiClient.ts for clarity.
* fix: add type and property checks for tool call handling
Improves robustness by adding explicit checks for 'function' property and 'type' when parsing tool calls and estimating tokens. Also adds error handling for unknown tool call types in mcp-tools and updates related test logic.
* feat(模型配置): 添加gpt5模型支持及相关配置
- 在模型类型中新增gpt5支持
- 添加gpt5系列模型检测函数
- 更新推理选项配置和国际化文本
- 调整effort ratio数值
* fix(ThinkingButton): 为gpt-5及后续模型添加minimal到low的选项回退映射
* feat(i18n): 更新思维链长度的中文翻译并调整对应图标
为思维链长度的"minimal"选项添加中文翻译"微念"，同时调整各选项对应的灯泡图标亮度
* feat(i18n): 为推理努力设置添加"minimal"选项并调整英文文案
* fix: openai patch
* wip: OpenAISettingsGroup display
* fix: 修复OpenAISettingsGroup组件中GPT5条件下的渲染逻辑
* refactor(OpenAISettingsGroup): 优化设置项的分组和分隔符逻辑
* feat(模型配置): 添加gpt-5到visionAllowedModels列表
* feat(模型配置): 添加gpt-5到函数调用支持列表
将gpt-5及其变体添加到FUNCTION_CALLING_MODELS支持列表，同时将gpt-5-chat添加到排除列表
* fix: 在OpenAI推理模型检查中添加gpt-5-chat支持
* Update OpenAISettingsGroup.tsx
* feat(模型支持): 添加对verbosity模型的支持判断
新增isSupportVerbosityModel函数用于判断是否支持verbosity模型
替换原有isGPT5SeriesModel判断逻辑，统一使用新函数
* fix: 修复支持详细程度模型的判断逻辑
使用 getLowerBaseModelName 处理模型 ID 以确保大小写不敏感的比较
* feat: 添加对gpt-5模型的网络搜索支持但不包括chat变体
* fix(models): 修复gpt5模型支持选项缺少'off'的问题
* fix: 添加gpt-5到支持Flex Service Tier的模型列表
* refactor(aiCore): 优化OpenAI verbosity类型定义和使用
移除OpenAIResponseAPIClient中冗余的OpenAIVerbosity导入
在BaseApiClient中明确getVerbosity返回类型为OpenAIVerbosity
简化OpenAIResponseAPIClient中verbosity的类型断言
* fix(openai): 仅在支持verbosity的模型中添加verbosity参数
* fix(i18n): 修正OpenAI设置中不一致的翻译
* fix: modify low effort ratio
* fix(openai): 修复GPT5系列模型在启用网页搜索时不能使用minimal reasoning_effort的问题
* fix(openai): 修复GPT5系列模型在启用web搜索时不能使用minimal推理的问题
---------
Co-authored-by: icarus <eurfelux@gmail.com>
Co-authored-by: Phantom <59059173+EurFelux@users.noreply.github.com>
Updated the logic for determining tool usage by consolidating conditions into a single variable. Improved the handling of MCP tools by filtering active servers and checking for enabled tools before sending notifications. Removed redundant code and ensured that tool execution completion notifications are sent based on the new logic.
* feat: add language filter to local web search queries
Integrates user's language setting into search queries by applying a language filter based on the provider (Google, Bing, or others). This improves search relevance for localized results.
* fix: update language filter logic in LocalSearchProvider
Refactored applyLanguageFilter to use provider.id instead of provider.name for search engine identification. Adjusted filtering for Google, Bing, and Baidu, and removed default language filter for other providers.
* Update LocalSearchProvider.ts
- Updated internal state synchronization for HTML content and preview.
- Enhanced preview refresh logic to check for content changes every 2 seconds.
- Improved comments for clarity and consistency in English.
- Added onChunkReceived call to fetchChatCompletion to improve response handling.
- Removed redundant onChunkReceived call from the completion parameters section in fetchChatCompletion.
refactor(ProxyManager): streamline proxy configuration and bypass rules handling
- Simplified proxy configuration logic in registerIpc by directly assigning bypass rules.
- Removed redundant bypass rules assignment in ProxyManager, initializing it as an empty array.
- Updated GeneralSettings to conditionally render based on custom proxy mode only.
- Removed unnecessary cursor styles from draggable items in the VirtualRow component.
- Updated snapshot tests to reflect changes in cursor styles.
- Introduced a new DragHandle component in ProviderSettings for improved drag interaction, with hover effects to enhance user experience.
Introduces logic to allow users to disable thinking mode directly from the ThinkingButton if supported. Tooltip text now reflects the action to close when thinking is enabled and 'off' is available.
* feat(discover): implement Discover feature with routing and UI components
- Added a new Discover page with sidebar and main content areas.
- Integrated routing for Discover, including subcategories and tabs.
- Created components for Discover sidebar and main content.
- Updated localization files to include new Discover titles and labels.
- Refactored existing components to accommodate the new Discover feature.
- Enhanced sidebar icons and navigation for better user experience.
* feat(discover): enhance Discover page with Tailwind CSS integration and routing improvements
- Added Tailwind CSS import to the entry point for styling.
- Updated the ThemeProvider to dynamically apply Tailwind themes based on user selection.
- Refactored Discover page to utilize new ROUTERS structure for better routing management.
- Simplified category handling in useDiscoverCategories hook by leveraging ROUTERS_ENTRIES.
- Introduced InternalCategory interface for better type management in Discover components.
- Cleaned up unused code and comments for improved readability.
* fix: update import statement for linguist-languages in update-languages.ts
* fix: standardize import quotes and improve localization files
- Updated import statements in use-mobile.ts and motionVariants.ts to use single quotes for consistency.
- Added new localization entries for the "discover" section in multiple language files, including English, Japanese, Russian, Traditional Chinese, Greek, Spanish, French, and Portuguese.
* refactor(discover): simplify Discover page structure and improve routing logic
- Refactored DiscoverPage component to streamline tab and sidebar handling.
- Updated routing logic to utilize a new ROUTERS_MAP for better category management.
- Removed unused props and simplified state management in useDiscoverCategories hook.
- Enhanced DiscoverSidebar and DiscoverMain components for improved clarity and performance.
- Adjusted CherryStoreType enum values for consistency in path definitions.
* fix: update file upload body type in MineruPreprocessProvider
- Changed the body of the fetch request from a Buffer to a Uint8Array to ensure proper handling of binary data during PDF uploads.
* fix: ensure Blob creation uses a copy of byte arrays for image handling
- Updated Blob creation in ImageGenerationMiddleware, ImageViewer, and MessageImage components to use `slice()` on byte arrays, preventing potential mutations of the original data.
* chore: update Vite React SWC plugin and adjust Electron config for conditional styling
- Upgraded `@vitejs/plugin-react-swc` from version 3.9.0 to 3.11.0 for improved performance and features.
- Modified Electron Vite configuration to conditionally apply styled-components plugin based on the VITEST environment variable.
- Updated snapshot tests for `InputEmbeddingDimension` and `Spinner` components to reflect style changes.
* chore: upgrade @swc/plugin-styled-components to version 9.0.2 in package.json and yarn.lock
* refactor: streamline styled-components plugin configuration in Electron Vite setup
- Consolidated the styled-components plugin configuration in the Electron Vite config file for improved readability and maintainability.
- Removed conditional application of the plugin based on the VITEST environment variable, ensuring consistent styling behavior across environments.
* i18n: update translations for discover section across multiple languages
- Replaced placeholder text with accurate translations for the "discover" section in English, Japanese, Russian, Traditional Chinese, Greek, Spanish, French, and Portuguese.
- Ensured consistency in terminology and improved clarity in user-facing messages.
* i18n: update "discover" title translations across multiple languages
- Updated the "discover" title in English, Japanese, Russian, Traditional Chinese, Greek, Spanish, French, and Portuguese to ensure accurate and consistent terminology.
- Adjusted related key mappings in the localization files for improved clarity in user-facing messages.
* chore: update lucide-react to version 0.536.0 and clean up tsconfig paths
* fix: update input style in snapshot tests and format message mentions in MessageContent component
* feat: disable mask closing for various popups across the application
- Updated multiple popup components to prevent closing when clicking outside, enhancing user experience and preventing accidental dismissals.
- Affected components include ImportAgentPopup, QuickPhrasesButton, NewAppButton, EditMcpJsonPopup, TopicNamingModalPopup, CustomHeaderPopup, and QuickPhraseSettings.
This change aims to improve the usability of modal dialogs by ensuring users must explicitly confirm or cancel their actions.
* feat: implement click outside to save edits in TopicsTab
- Added a useEffect hook to handle clicks outside the editing input, triggering save on blur.
- Updated onClick behavior for topic items to prevent switching while editing.
- Enhanced cursor style for better user experience during editing.
This change improves the editing experience by ensuring that edits are saved when the user clicks outside the input field.
* feat: integrate in-place editing for topic names in TopicsTab
- Added useInPlaceEdit hook to manage topic name editing, improving user experience.
- Removed previous editing logic and integrated new editing flow with save and cancel functionalities.
- Updated UI interactions to reflect the new editing state, ensuring a smoother editing process.
This change enhances the editing experience by allowing users to edit topic names directly within the list, streamlining the workflow.
* feat: add OpenAI o3 model support with enhanced tool calling
- Add o3 and o3-mini model definitions with reasoning effort support
- Implement o3-compatible strict schema validation for MCP tools
- Add comprehensive o3 schema processing with DRY improvements
- Extract reusable schema processing functions for maintainability
- Add 15+ test cases validating o3 strict mode requirements
- Fix schema composition keyword handling with loop-based approach
- Ensure ALL object schemas have complete required arrays for o3
- Support tool calling with proper o3 schema transformations
This enables OpenAI o3 models to work properly with MCP tool calling
while improving code organization and test coverage.
Signed-off-by: Luke Galea <luke@ideaforge.org>
* Remove redundant reference in HtmlArtifactsPopup.tsx
* refactor: move filterProperties to mcp-schema, fix tests
---------
Signed-off-by: Luke Galea <luke@ideaforge.org>
Co-authored-by: one <wangan.cs@gmail.com>
Co-authored-by: suyao <sy20010504@gmail.com>
* fix[AgentsPage]: fix using name deduplication leads to the loss of agents with the same name
* fix[agents-zh]: fix id 499 the problem of markdown display
* fix[agent]: agent search adds descriptive text search
* feat(ProxyManager): implement SelectiveDispatcher for localhost handling
- Added SelectiveDispatcher to manage proxy and direct connections based on the hostname.
- Introduced isLocalhost function to check for localhost addresses.
- Updated ProxyManager to bypass proxy for localhost in dispatch methods and set proxy bypass rules.
- Enhanced global dispatcher setup to utilize SelectiveDispatcher for both EnvHttpProxyAgent and SOCKS dispatcher.
* refactor(ProxyManager): update axios configuration to use fetch adapter
- Changed axios to use the 'fetch' adapter for proxy requests.
- Removed previous proxy settings for axios, streamlining the configuration.
- Updated HTTP methods to bind with the new proxy agent.
* feat(Proxy): add support for proxy bypass rules
- Updated IPC handler to accept optional bypass rules for proxy configuration.
- Enhanced ProxyManager to store and utilize bypass rules for localhost and other specified addresses.
- Modified settings and UI components to allow users to input and manage bypass rules.
- Added translations for bypass rules in multiple languages.
* feat(ProxyManager): add HTTP_PROXY environment variable support
- Added support for the HTTP_PROXY environment variable in ProxyManager to enhance proxy configuration capabilities.
* lint
* refactor(ProxyManager): optimize bypass rules handling
- Updated bypass rules initialization to split the rules string into an array for improved performance.
- Simplified the isByPass function to directly check against the array of bypass rules.
- Enhanced configuration handling to ensure bypass rules are correctly parsed from the provided settings.
* refactor(ProxyManager): streamline bypass rules initialization
- Consolidated the initialization of bypass rules by directly splitting the default rules string into an array.
- Updated configuration handling to ensure bypass rules are correctly assigned without redundant splitting.
* style(GeneralSettings): adjust proxy bypass rules input width to improve UI layout
* refactor(ProxyManager): enhance proxy configuration logging and handling
- Added proxy bypass rules to the configuration method for improved flexibility.
- Updated logging to include bypass rules for better debugging.
- Refactored the setGlobalProxy method to accept configuration parameters directly, streamlining proxy setup.
- Adjusted the useAppInit hook to handle proxy settings more cleanly.
* refactor(ProxyManager): implement close and destroy methods for proxy dispatcher
- Added close and destroy methods to the SelectiveDispatcher class for better resource management.
- Updated ProxyManager to handle the lifecycle of the proxyDispatcher, ensuring proper closure and destruction.
- Enhanced error handling during dispatcher closure and destruction to prevent resource leaks.
* refactor(ProxyManager): manage proxy agent lifecycle
- Introduced proxyAgent property to ProxyManager for better management of the proxy agent.
- Implemented error handling during the destruction of the proxy agent to prevent potential issues.
- Updated the proxy setup process to ensure the proxy agent is correctly initialized and cleaned up.
* refactor(ProxyManager): centralize default bypass rules management
- Moved default bypass rules to a shared constant for consistency across components.
- Updated ProxyManager and GeneralSettings to utilize the centralized bypass rules.
- Adjusted migration logic to set default bypass rules from the shared constant, ensuring uniformity in configuration.
* refactor: align lucide icons in antd button
* refactor(AssistantsTab): use lucide icon and typography in add assistant button
* refactor: use lucide icons for assistant item dropdown
* refactor: use lucide icons in topic item dropdown
* refactor: use lucide icon in InfoTooltip, align ApiOptionsSettings expand icon
* refactor: use lucide icons in TokenCount
* refactor: use brush in assistant item dropdown
* test: update snapshot
* test: mock tooltip
* fix: token count alignment
* refactor: update icons in MessageMenubar, bump antd
* refactor: use lucide icons in MessageTools, make colors consistent
* refactor: use lucide icons in ProviderSetting
* test: simplify test with mocks
* refactor: use lucide icons in knowledge base dropdown
* refactor: export all custom icons, use EditIcon for lucide pen
* refactor: use lucide copy for CopyIcon, update tests
* refactor: use lucide icons in MessageMenubar
* refactor: improve pause and send button style
* refactor: export SvgSpinners180Ring as LoadingIcon
* refactor: use lucide icons in Agents, use DeleteIcon
* refactor: use Pencil as EditIcon
* fix: i18n key missing
* refactor: use lucide icons in Files
* refactor: use lucide icons in KnowledgeBase items
* refactor: use lucide icons in assistant settings
* refactor: use lucide icons in memory settings, add UserSelector
* chore: remove duplicate memory component
* refactor: use lucide icons in ProviderList
* refactor: use lucide icons in QuickPhraseSettings
* refactor: use lucide icons in McpSettings
* refactor: use lucide icons in DataSettings
* refactor: use lucide icons in DefaultAssistantSettings
* refactor: add icon to save
* refactor: add lucide-custom
* fix: icon position in ModelEditContent
* refactor: use ListMinus in ManageModelsList
* refactor: improve TokenCount alignment
* fix: topic pin/unpin i18n
* fix: self review
* fix: simplify knowledge base dropdown
* fix: remove plus icon color
* refactor: add ResetIcon and RefreshIcon
* refactor: update theme handling in TabContainer and Sidebar components
* refactor: replace setTheme with toggleTheme and update theme state management
* feat: add Monitor icon for additional theme state in both components
* feat(DisplaySettings): replace theme icons with lucide icons for improved visual representation
* updated light and dark theme icons to use Sun and Moon from lucide
* replaced SyncOutlined with Monitor icon for system theme option
* format code
* feat(TabContainer): add tooltip for theme toggle button with translation support
* integrated Tooltip component to provide contextual information for the theme toggle button
* utilized useTranslation hook for internationalization of tooltip text
* imported getThemeModeLabel to enhance theme description
- use DynamicVirtualList in SelectModelPopup
- use DynamicVirtualList in QuickPanelView
- remove react-window
- simplify SelectModelPopup states, improve maintainability
* perf: inlining completeLineTokens and use memo for minor improvements
* chore: bump shiki to 3.9.1
* refactor: improve token line
* refactor: add plainTokenStyle
* feat(database): enable strict transaction durability for CherryStudio database
- Updated the Dexie database initialization to include `chromeTransactionDurability: 'strict'`, enhancing data integrity during transactions.
* feat(app): enhance application shutdown process and data flushing
- Added functionality to flush storage data and cookies before quitting the application, ensuring data integrity.
- Introduced a new `handleBeforeQuit` function to centralize cleanup logic for both manual and update-triggered quits.
- Updated logging to provide better insights during the shutdown process.
- Modified ProxyManager to use debug level for unchanged proxy configurations.
- Added `persistor` to the global window object and implemented `handleSaveData` to flush Redux state before quitting.
* format code
* feat(ipc): add App_SaveData channel and implement data saving on window close
- Introduced a new IPC channel `App_SaveData` for saving application data.
- Updated `WindowService` to send a save data message when the main window is closed.
- Enhanced `useAppInit` hook to handle the `App_SaveData` event and trigger data saving logic.
* refactor(env): remove persistor from global window object
- Removed the `persistor` property from the global `window` object in both `env.d.ts` and `index.ts` files, streamlining the application state management.
- add a custom dynamic virtual list component
- add tests
- support autohide
- used in ManageModelsList, ModelListGroup, KnowledgePage, FileList
- improve DraggableVirtualList
- use name DraggableVirtualList directly, make it flex by default
- use DraggableVirtualList in ProviderList
- Introduced a new ApiServerSettings component for managing API server configurations.
- Updated ToolSettings to include API server options and controls.
- Enhanced GeneralSettings to improve proxy settings management.
- Refactored UI elements for better organization and user experience.
* fix(prompt): fix variable replacement in function mode
* fix(i18n): update available variables in prompt tip
* feat(prompt): replace prompt variable in Prompt and AssistantPromptSettings components
* fix(prompt): add fallback value if replace failed
* feat(prompt): add hook and settings for automatic prompt replacement
* feat(prompt): add supported variables and utility function to check if they exist
* feat(prompt): enhance variable handling in prompt settings and tooltips
* feat(i18n): add prompt settings translations for multiple languages
* refactor(prompt): remove debug log from prompt processing
* fix(prompt): handle model name variables and improve prompt processing
* fix: correct variable replacement setting and update migration defaults
* remove prompt settings
* refactor: simplify model name replacement logic
- Remove unnecessary assistant parameter from buildSystemPrompt function
- Update all API clients to use the simplified function signature
- Centralize model name replacement logic in promptVariableReplacer
- Improve code maintainability by reducing parameter coupling
* fix: eslint error
* refactor: remove unused interval handling in usePromptProcessor
* test: add tests, remove redundant replacing
* feat: animate prompt substitution
* chore: prepare for merge
* refactor: update utils
* refactor: remove getStoreSettings
* refactor: update utils
* style(Message/Prompt): 禁止文本选中以提升用户体验
* fix(Prompt): 修复内存泄漏问题，清除内部定时器
* refactor: move prompt replacement to api service
---------
Co-authored-by: one <wangan.cs@gmail.com>
Co-authored-by: icarus <eurfelux@gmail.com>
* ✨ feat: make API server default to closed/disabled
- Add `enabled` field to ApiServerConfig interface with default value false
- Update Redux store to include apiServer.enabled setting (defaults to false)
- Add setApiServerEnabled action to control server state
- Modify main process to only auto-start API server if explicitly enabled
- Update UI to show enable/disable toggle in API server settings
- Add migrations to handle existing configurations
- Server controls now only visible when API server is enabled
The API server will no longer start automatically on application launch
unless explicitly enabled by the user through the settings interface.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
* 🔧 chore: improve build and lint commands
- Move typecheck and i18n checks to lint command for better developer experience
- Simplify build:check to focus on linting and testing
- Ensure all quality checks run together during development
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
---------
Co-authored-by: Claude <noreply@anthropic.com>
* refactor(ApiService): optimize memory search handling and improve error logging
- Updated fetchExternalTool to handle memory search in a separate promise, ensuring better management of asynchronous operations.
- Enhanced error logging for memory search failures to improve debugging.
- Removed redundant memory search result storage logic from the main result handling section.
* refactor(ApiService): streamline external tool fetching and memory search handling
- Simplified the fetchExternalTool function by separating web and knowledge search calls into individual await statements.
- Improved memory search handling by awaiting the searchMemory function directly, ensuring better asynchronous flow.
- Updated result storage logic to ensure memory search results are stored correctly after fetching.
- improve responsiveness for provider switching
- improve ModelList performance
- use startTransition
- use virtual list for model groups
- remove excessive ModelEditContent, add EditModelPopup
- add model count
- improve model management popup
- rename to ManageModelsPopup
- use virtual list with sticky group name
- use SvgSpinners180Ring for spin
- Added "Start" and "Stop" labels in English, Japanese, Russian, Chinese (Simplified and Traditional) locales.
- Introduced "Authorization Header" title and descriptions for API key and port fields across all locales.
- Removed deprecated API documentation unavailable messages for a cleaner user experience.
- Added a new handleTabClick function to streamline tab navigation and hide the minapp popup when a tab is clicked.
- Implemented an animation for the minapp switch indicator in the TopNavbarOpenedMinappTabs component, improving visual feedback during tab switching.
- Refactored the rendering of minapp items to use Tooltip for better accessibility and user experience.
- Removed unnecessary StyledLink components to simplify the structure of the navigation items.
* fix(AppUpdater): simplify error logging and update version check logic
- Updated error logging to use a more concise format.
- Changed logging messages for update events to be more consistent.
- Modified the update check logic to return null when no update is available.
- Enhanced the app initialization hook to include update availability in the state dispatch.
* fix(useAppInit): simplify update state dispatch logic by removing update availability check
- Introduced a new function to check if a model is the Qwen 3-235B-A22B thinking model.
- Updated the ThinkingButton component to utilize the new detection function for improved model handling.
* refactor(ModelEditContent): enhance model capabilities management and introduce uniqueObjectArray utility
- Updated ModelEditContent to improve handling of model capabilities, ensuring user selections are accurately reflected.
- Introduced a new utility function, uniqueObjectArray, to filter out duplicate objects in arrays, enhancing data integrity.
- Refactored logic for updating model capabilities to utilize the new utility, streamlining the process and improving code clarity.
* refactor(ModelEditContent): enhance model capabilities management with useEffect and improved type handling
- Added useEffect to manage model capabilities based on user selections and showMoreSettings state.
- Refactored logic to streamline the handling of default and selected model types, improving clarity and maintainability.
- Utilized useRef to track changed types, ensuring accurate updates to model capabilities during user interactions.
* refactor(ModelEditContent): optimize model capabilities update logic with getUnion utility
- Enhanced the model capabilities management by integrating the getUnion utility to streamline the merging of selected types and unselected capabilities.
- Improved clarity in the useEffect hook for managing model capabilities based on user selections and the showMoreSettings state.
- Refactored condition checks for updating user selections to ensure accurate handling of model capabilities during interactions.
* refactor(ModelEditContent): improve model capabilities reset logic and enhance debugging
- Introduced a cloneDeep utility to preserve original model capabilities for reset functionality.
- Updated the handleResetTypes function to restore original capabilities instead of clearing them.
- Added console logs for better debugging and tracking of model capabilities during updates.
* feat(ModelEditContent): track user modifications for model capabilities
- Added a state variable to track if the user has modified model capabilities.
- Updated the handleTypeChange function to set the modification flag when types are changed.
- Modified the reset button to only display when there are user modifications, enhancing the user interface and interaction clarity.
* feat(models): add support for Zhipu model and enhance reasoning checks
- Introduced support for the new Zhipu model (GLM-4.5) in the models configuration.
- Added functions to check if the Zhipu model supports reasoning and thinking tokens.
- Updated the ThinkingButton component to recognize Zhipu model options and integrate them into the reasoning effort logic.
- Ensured that the reasoning checks are streamlined to include the new Zhipu model alongside existing models.
* feat(models): expand Zhipu model support and refine reasoning checks
- Added new Zhipu models: GLM-4.5-Flash, GLM-4.5-AIR, GLM-4.5-AIRX, and GLM-4.5-X to the models configuration.
- Enhanced the isZhipuReasoningModel function to include checks for the new models and ensure robust reasoning validation.
- Removed redundant checks for 'glm-z1' from the isReasoningModel function to streamline logic.
* fix(LocalBackup): streamline local backup directory handling
- Added a new state to manage the resolved local backup directory in LocalBackupSettings.
- Updated the LocalBackupManager component to use the resolved directory instead of the raw input.
- Enhanced the backupToLocal and restoreFromLocal functions to resolve the local backup directory path before use, improving reliability.
* refactor(Backup): remove setLocalBackupDir functionality
- Removed the setLocalBackupDir method and its associated IPC channel handling from the BackupManager and IpcChannel.
- Updated LocalBackupSettings to eliminate direct calls to setLocalBackupDir, instead resolving the local backup directory path directly.
- Cleaned up related code in the BackupService to streamline local backup operations.
* format code
* feat: add Top-P parameter toggle with default enabled state and improved UI styling
* fix: resolve undefined enableTopP issue in ppio models by using getAssistantSettings
* refactor(api): Unify getTopP method to BaseApiClient
* feat(settings): adjust layout of Top-P setting in assistant model settings
* feat: add temperature parameter toggle control with UI and multi-language support
* fix: Fix lint error
* fix: Sort these imports
* style(settings): refactor model settings layout and styles
* chore: yarn sync:i18n
* feat(ipc): add isPathInside functionality to check path relationships
- Introduced a new IPC channel for checking if a path is inside another path, enhancing path validation capabilities.
- Implemented the isPathInside function in the file utility, which accurately determines parent-child path relationships, handling edge cases.
- Updated relevant components to utilize the new isPathInside function for validating app data and backup paths, ensuring better user experience and error handling.
- Added comprehensive tests for isPathInside to cover various scenarios, including edge cases and error handling.
* format code
- Removed the specific check for Grok models in OpenAIApiClient and consolidated it with the general reasoning effort model check.
- Added support for a new Perplexity model, 'sonar-deep-research', in the models configuration.
- Updated the reasoning model checks to include Perplexity models in the models.ts file.
- Enhanced the ThinkingButton component to recognize and handle Perplexity model options.
* feat(ModelList): add support for 'supported_text_delta' in model configuration
- Introduced a new boolean property 'supported_text_delta' to the model configuration, allowing models to indicate support for text delta outputs.
- Updated the AddModelPopup and ModelEditContent components to handle this new property, including UI elements for user interaction.
- Enhanced migration logic to set default values for existing models based on compatibility checks.
- Added corresponding translations for the new property in the i18n files.
* feat(OpenAIApiClient): enhance support for Qwen MT model and system message handling
- Added support for Qwen MT model in OpenAIApiClient, including translation options based on target language.
- Updated system message handling to accommodate models that do not support system messages.
- Introduced utility functions to identify Qwen MT models and their compatibility with text delta outputs.
- Enhanced TextChunkMiddleware to handle text accumulation based on model capabilities.
- Updated model configuration to include Qwen MT in the list of excluded models for function calling.
* feat(i18n): add translations for 'supported_text_delta' in multiple languages
- Introduced new translation entries for the 'supported_text_delta' property in English, Japanese, Russian, and Traditional Chinese localization files.
- Updated the corresponding labels and tooltips to enhance user experience across different languages.
* refactor(ModelEditContent): reposition 'supported_text_delta' input for improved UI layout
- Moved the 'supported_text_delta' Form.Item from its previous location to enhance the user interface and maintain consistency in the model editing experience.
- Ensured that the switch input for 'supported_text_delta' is now displayed in a more logical order within the form.
* fix(TextChunkMiddleware): update condition for supported_text_delta check
- Changed the condition in TextChunkMiddleware to explicitly check for 'supported_text_delta' being false, improving clarity in the logic.
- Updated test cases to reflect the new structure of model configurations, ensuring 'supported_text_delta' is consistently set to true for relevant models.
* feat(migrate): add support for 'supported_text_delta' in assistant models
- Updated migration logic to set 'supported_text_delta' for both default and specific models within assistants.
- Implemented checks to ensure compatibility with text delta outputs, enhancing model configuration consistency.
* feat(ModelList): add 'supported_text_delta' to model addition logic
- Enhanced the model addition process in EditModelsPopup, NewApiAddModelPopup, and NewApiBatchAddModelPopup to include the 'supported_text_delta' property.
- Ensured consistency across components by setting 'supported_text_delta' to true when adding models, improving compatibility with text delta outputs.
* feat(migrate): streamline model text delta support in migration logic
- Refactored migration logic to utilize a new helper function for updating the 'supported_text_delta' property across various model types, enhancing code clarity and reducing redundancy.
- Ensured that all relevant models, including those in assistants and LLM providers, are correctly configured for text delta compatibility during migration.
* feat(OpenAIApiClient): integrate language mapping for Qwen MT model translations
- Updated the OpenAIApiClient to utilize a new utility function for mapping target languages to Qwen MT model names, enhancing translation accuracy.
- Refactored migration logic to ensure default tool use mode is set for assistants lacking this configuration, improving user experience.
- Added a new utility function for language mapping in the utils module, supporting better integration with translation features.
* feat(ModelList): update model addition logic to determine 'supported_text_delta'
- Integrated a new utility function to assess model compatibility with text delta outputs during the model addition process in AddModelPopup.
- Simplified the logic for setting the 'supported_text_delta' property, enhancing clarity and ensuring accurate model configuration.
* feat(ModelList): unify model addition logic for 'supported_text_delta'
- Refactored model addition across AddModelPopup, EditModelsPopup, NewApiAddModelPopup, and NewApiBatchAddModelPopup to consistently determine 'supported_text_delta' using a utility function.
- Simplified the logic for setting 'supported_text_delta', enhancing clarity and ensuring accurate model configuration across components.
* feat(models): add support for new Qwen model and update effort ratio
- Introduced support for the 'qwen3-235b-a22b-thinking' model with a max token limit of 81,920.
- Updated the model options in ThinkingButton to include 'qwen_3235ba22b_thinking'.
- Adjusted the effort ratio for 'low' from 0.2 to 0.05 for improved performance.
* chore: remove
* feat(MCPService, MCPSettings, MessageTools): enhance long-running server support and UI integration
- Added support for long-running server configurations in MCPService, allowing for timeout adjustments based on server settings.
- Introduced a new `longRunning` property in MCPSettings to manage server behavior and UI elements accordingly.
- Integrated a ProgressBar component in MessageTools to visually represent progress for long-running operations, improving user experience.
* refactor(IpcChannel, MCPService, MessageTools): remove progress IPC channel and integrate progress handling
- Removed the `Mcp_SetProgress` channel from `IpcChannel` and its associated handlers in `ipc.ts` and `preload/index.ts`.
- Integrated progress handling directly in `McpService` to send progress updates to the main window.
- Updated `MessageTools` to display progress using Ant Design's `Progress` component, enhancing the user interface for long-running operations.
- Deleted the `ProgressBar` component as its functionality has been replaced by the new progress handling approach.
* feat(MCPService): add maxTotalTimeout configuration for long-running operations
- Introduced a new `maxTotalTimeout` property in MCPService to define a maximum timeout duration for long-running server operations, enhancing control over server behavior based on the `longRunning` setting.
* refactor(MCPService): remove unused notification handler for cancelled operations
* Removed the CancelledNotificationHandler from MCPService to streamline notification handling and improve code clarity.
* Updated MessageTools component to simplify rendering logic for status indicators, enhancing readability and maintainability.
* refactor(knowledge): simplify dimension settings, support base migration
Embedding dimension
- remove 'auto set dimension', let the user take control
- reuse findModelById
- improve error messages for VoyageEmbeddings
Knowledgebase migration
- enable migration when model or dimension changes
- add knowledgeThunk to reuse code
KnowledgeSettings
- unify UI for AddKnowledgeBasePopup and EditKnowledgeBasePopup
- refactor KnowledgeSettings, split it to smaller components
Tests:
- knowledgeThunk
- InputEmbeddingDimension
- KnowledgeBaseFormModal
- GeneralSettingsPanel
- AdvancedSettingsPanel
- InfoTooltip
Misc.
- add InfoTooltip
- remove MemoriesSettingsModal
* fix: i18n and vitest config
* refactor(useSmoothStream): optimize chunk handling and state management
- Replaced state management with refs for chunk queue to improve performance and reduce unnecessary re-renders.
- Introduced Intl.Segmenter for better character segmentation based on language support.
- Updated rendering logic to ensure final text is displayed correctly when the stream ends.
- Cleaned up unused effects and comments for improved code clarity.
* refactor(useSmoothStream): move segmenter initialization outside of addChunk function
- Moved the initialization of Intl.Segmenter to the top of the file for better performance and to avoid redundant instantiation within the addChunk callback.
- This change enhances the efficiency of chunk processing in the useSmoothStream hook.
Use spread operator to copy all fields from parsed JSON instead of manually
mapping specific fields. This ensures all MCPServer properties including
headers, timeout, dxtVersion, and other optional fields are preserved during
JSON import.
Previously missing fields:
- headers: Record<string, string>
- timeout: number
- dxtVersion: string
- dxtPath: string
- Other optional MCPServer properties
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-authored-by: Claude <noreply@anthropic.com>
* chore(env): add .env.example file and update .gitignore
- Introduced a new .env.example file with NODE_OPTIONS configuration.
- Updated .gitignore to exclude .env.example from being ignored.
- Added instructions in dev.md for copying .env.example to .env.
* fix(MessageTools): improve error handling and logging in message preview rendering (#8453)
- Enhanced the rendering logic for message previews by adding a try-catch block to handle JSON parsing errors more gracefully.
- Updated the error handling to provide clearer error messages in the preview when exceptions occur.
- Added debug logging to track the rendering process of message content.
* refactor(Theme): update theme management to use setTheme function
- Replaced toggleTheme with setTheme for more explicit theme handling.
- Removed unused SunMoon icon from TabContainer and Sidebar components.
- Updated theme icon rendering logic to directly reflect the current theme state.
- Adjusted ThemeProvider to include setTheme in context for better theme management.
* refactor(ModelList): streamline button layout and improve accessibility
- Removed tooltip wrappers from manage and add model buttons for a cleaner UI.
- Introduced a new Flex container for primary and default buttons, enhancing layout consistency.
- Updated button rendering to improve accessibility and user experience.
* feat(ModelList): add bulk add/remove functionality for models with confirmation dialog
- Implemented onAddAll and onRemoveAll functions to handle bulk actions for models.
- Added confirmation dialog for adding all models to the list, enhancing user experience.
- Updated translations for confirmation messages in multiple languages.
* chore(languages): update languages with a script (#8445)
* chore(languages): update languages with a script
* refactor: update languages and merge it into constants
* refactor: add usf and ush
* refactor(ipc): enhance write permission check and add untildify utility
- Updated the hasWritePermission function to resolve paths using the new untildify utility, improving path handling.
- Modified IPC handler to await the permission check for better asynchronous handling.
- Introduced a new untildify function to convert paths starting with '~' to the user's home directory.
* fix(ModelEdit): enhance model type management and introduce new selection logic (#8420)
* fix(ModelEdit): enhance model type management and introduce new selection logic
- Added support for 'rerank' model type in the ModelEditContent component.
- Refactored type selection logic to utilize new utility functions for finding differences and unions in model types.
- Updated model type handling to include user selection status, improving user experience in type management.
- Adjusted migration logic to initialize newType for existing models, ensuring backward compatibility.
- Introduced isUserSelectedModelType utility to streamline model type checks across the application.
* refactor(isFunctionCallingModel): simplify model type check logic
- Replaced the inline check for 'function_calling' model type with a call to the new utility function isUserSelectedModelType, enhancing code clarity and maintainability.
* feat(collection): add utility functions for array operations
- Introduced `findIntersection`, `findDifference`, and `findUnion` functions to handle array operations with support for custom key selectors and comparison functions.
- Removed previous implementations from `index.ts` to streamline utility exports.
- Added comprehensive tests for new functions covering basic types and object types with various edge cases.
* refactor(collection): rename utility functions for clarity
- Renamed `findIntersection`, `findDifference`, and `findUnion` to `getIntersection`, `getDifference`, and `getUnion` respectively for improved clarity and consistency in naming.
- Updated corresponding tests to reflect the new function names, ensuring all tests pass with the updated utility functions.
* refactor(ModelEditContent): update model type management and improve selection logic
- Replaced utility function calls to `findDifference` and `findUnion` with `getDifference` and `getUnion` for consistency.
- Introduced temporary state management for model types to enhance user selection handling.
- Added a reset functionality for model type selections, improving user experience.
- Updated the rendering logic to conditionally disable certain model types based on user selections.
* fix(ModelEditContent): enhance model type selection logic with conditional disabling
- Introduced logic to conditionally disable 'rerank' and 'embedding' model types based on user selections.
- Updated the state management for model types to ensure correct user selection handling.
- Improved the confirmation modal to reflect the updated selection logic for better user experience.
* fix(ModelEditContent): refine model type selection and update confirmation logic
- Enhanced the logic for model type selection to ensure accurate user selections for 'rerank' and 'embedding'.
- Updated the confirmation modal to reflect changes in selection handling, improving user experience.
- Adjusted state management to correctly handle updates based on selected model types.
* fix(models): update model support logic to include 'qwen3-235b-a22b-instruct'
* refactor(models): rename 'newType' to 'capabilities' and update related logic in ModelEditContent and migration scripts
* feat(ipc): add App_ResolvePath channel and update path handling
- Introduced a new IPC channel `App_ResolvePath` to resolve file paths, enhancing path management.
- Updated the `hasWritePermission` function to log the original directory instead of the resolved one.
- Modified the `LocalBackupSettings` component to utilize the new `resolvePath` method for improved directory validation.
* add ut
* fix comments
* fix clear manually
* delete duplicate var
---------
Co-authored-by: kangfenmao <kangfenmao@qq.com>
Co-authored-by: SuYao <sy20010504@gmail.com>
Co-authored-by: one <wangan.cs@gmail.com>
* refactor(CodeEditor): add fontSize to props
* feat: use CodeEditor in assistant prompt settings
* refactor(font): add code font family for windows
* refactor: support Sarasa
- Removed the reasoning check from the ThinkChunkMiddleware to streamline the processing logic.
- This change enhances the middleware's efficiency by focusing on handling streams directly without the reasoning condition.
- Updated navigation logic in `useMCPServers` and `McpServersList` to use server IDs in the URL for better routing.
- Implemented scroll position memory in `McpServersList` to enhance user experience when navigating back to the server list.
- Adjusted route parameters in `MCPSettings` to retrieve server IDs from the URL, improving data handling and clarity.
* fix(mcp-tools): enhance tool lookup logic to support partial matches
- Updated the tool lookup logic in `geminiFunctionCallToMcpTool` to include partial matches for both tool IDs and names, improving the flexibility of tool identification.
* refactor(mcp-tools): simplify tool lookup logic for improved clarity
- Refactored the tool lookup logic in `geminiFunctionCallToMcpTool` to streamline the identification process by consolidating checks for tool IDs and names into a single variable. This enhances readability and maintains functionality for partial matches.
* feat(migrate): add default settings for assistants during migration
- Introduced a new migration step to assign default settings for assistants that lack configuration.
- Default settings include temperature, context count, and other parameters to ensure consistent behavior across the application.
* chore(store): increment version number to 115 for persisted reducer
* feat(vertex-sdk): integrate Anthropic Vertex SDK and add access token retrieval
- Added support for the new `@anthropic-ai/vertex-sdk` in the project.
- Introduced a new IPC channel `VertexAI_GetAccessToken` to retrieve access tokens.
- Implemented `getAccessToken` method in `VertexAIService` to handle service account authentication.
- Updated the `IpcChannel` enum and related IPC handlers to support the new functionality.
- Enhanced the `VertexAPIClient` to utilize the `AnthropicVertexClient` for model handling.
- Refactored existing code to accommodate the integration of the Vertex SDK and improve modularity.
* feat(vertex-ai): enhance VertexAI settings and API host management
- Added a new method to format the API host URL in both AnthropicVertexClient and VertexAPIClient.
- Updated getBaseURL methods to utilize the new formatting logic.
- Enhanced VertexAISettings component to include an input for API host configuration, with help text for user guidance.
- Updated localization files to include new help text for the API host field in multiple languages.
* fix(vertex-sdk): update baseURL handling and patch dependencies
- Refactored baseURL assignment in AnthropicVertexClient to ensure it defaults to undefined when the URL is empty.
- Updated yarn.lock to reflect changes in dependency resolution and checksum for @anthropic-ai/vertex-sdk patch.
* refactor(VertexAISetting): use provider.id rather than provider
* refactor: improve API host formatting in AnthropicVertexClient
- Updated the `formatApiHost` method to streamline host URL handling.
- Introduced a helper function to determine if the original host should be used based on its format.
- Ensured consistent appending of the `/v1/` path for valid API requests.
* fix: handle empty host in AnthropicVertexClient
- Added a check in the `getBaseURL` method to return the host if it is empty, preventing potential errors.
- Included a console log for the base URL to aid in debugging and verification of the URL formatting.
* feat(AnthropicVertexClient): add logging for authentication errors and mock client in tests
- Introduced logging functionality in AnthropicVertexClient to replace console.error with logger service for better error tracking.
- Added mock implementation for AnthropicVertexClient in tests to enhance testing capabilities.
- Updated package.json to include the @aws-sdk/client-s3 dependency.
* feat(tests): add comprehensive tests for client compatibility types
- Introduced a new test file to validate compatibility types for various API clients including OpenAI, Anthropic, Gemini, Aihubmix, NewAPI, and Vertex.
- Implemented mock services to facilitate testing and ensure isolation of client behavior.
- Added tests for both direct API clients and decorator pattern clients, ensuring correct compatibility type returns.
- Enhanced middleware compatibility logic tests to verify correct identification of compatible clients.
---------
Co-authored-by: one <wangan.cs@gmail.com>
- Renamed `tempModelTypes` to `modelCapabilities` for improved clarity in the ModelEditContent component.
- Updated state management and logic to consistently use the new `modelCapabilities` variable throughout the component, enhancing readability and maintainability.
- Increased padding for the AppsContainerWrapper to 50px for better spacing.
- Added conditional padding for when the navbar is positioned at the top, reverting to 20px for improved layout consistency.
- Removed unnecessary useCallback functions for toggling assistants and topics, directly using the toggle functions instead.
- Eliminated the unused fullscreen hook import to clean up the code.
- Updated click handlers in the Navbar components for better readability and efficiency.
* fix(RawStreamListenerMiddleware): update model check for Anthropic API integration
- Replaced provider type check with model ID check to enhance compatibility with Claude models.
- Improved clarity in the middleware logic for handling raw output from the SDK.
* refactor(RawStreamListenerMiddleware): enhance model identification for Anthropic integration
- Introduced a new utility function `isAnthropicModel` to streamline model checks across the codebase.
- Updated middleware logic to utilize the new function for improved clarity and maintainability.
- Adjusted related tests to ensure compatibility with the updated model identification approach.
* test(ApiService.test): add mock for isAnthropicModel to enhance test coverage for model identification
* feat(ThinkingTagExtraction): add new tag configuration for 'kimi-vl-a3b-thinking' model and update model regex patterns in config
- Introduced a new tag configuration for the 'kimi-vl-a3b-thinking' model in ThinkingTagExtractionMiddleware.
- Updated models.ts to include regex patterns for 'kimi-vl-a3b-thinking', 'llama-guard-4', and 'llama-4' to enhance model compatibility.
* feat(models): add regex pattern for 'gemma3' model to enhance model compatibility
