* refactor(OpenAIClient): improve error handling in chat completion
Improve error handling in the chat completion process to handle scenarios where the response has no choices or the message is undefined. This ensures that the intermediate reply is returned in such cases. Also, add logging statements for better debugging.
* refactor(Markdown): only show blinking cursor for latest message
* refactor: add submitting selector to prevent blinking cursor for empty, non-streaming, non-latest messages
* refactor: use parseCompactConvo in buildOptions, and generate no default values for the API to avoid weird model behavior with defaults
* refactor: OTHER - always show cursor when markdown component is empty (preferable to not)
* refactor(OpenAISettings): use config object for setting defaults app-wide
* refactor: Use removeNullishValues in buildOptions for ALL endpoints
* fix: add missing conversationId to title methods for transactions; refactor(GoogleClient): model options, set no default, add todo note for recording token usage
* fix: at minimum set a model default, as is required by API (edge case)
* feat: basic invite-user script
* feat: add invite user functionality and registration validation middleware
* fix: invite user fixes
* refactor: consolidate direct model access to a central place of functions
* style(Registration): add spinner to continue button
* refactor: import ordrer
* feat: improve invite user script and error handling
* fix: merge conflict
* refactor: remove `console.log` and use `logger`
* fix: token operation and checkinvite issues
* bring back comment and remove console log
* fix: return invalid token when token is not found
* fix: getInvite fix
* refactor: Update Token.js to use async/await syntax for update and delete operations
* feat: Refactor Token.js to use async/await syntax for createToken and findToken functions
* refactor(inviteUser): define functions outside of module.exports
* Update AuthService.js
---------
Co-authored-by: Danny Avila <danny@librechat.ai>
* wip: initial cache control implementation, add typing for transactions handling
* feat: first pass of Anthropic Prompt Caching
* feat: standardize stream usage as pass in when calculating token counts
* feat: Add getCacheMultiplier function to calculate cache multiplier for different valueKeys and cacheTypes
* chore: imports order
* refactor: token usage recording in AnthropicClient, no need to "correct" as we have the correct amount
* feat: more accurate token counting using stream usage data
* feat: Improve token counting accuracy with stream usage data
* refactor: ensure more accurate than not token estimations if custom instructions or files are not being resent with every request
* refactor: cleanup updateUserMessageTokenCount to allow transactions to be as accurate as possible even if we shouldn't update user message token counts
* ci: fix tests
* feat: new dropdown
* fix: maintain popover active when open
* fix: update DeleteButton and ShareButton component to use useState for managing dialog state
* BREAKING: style improvement of base Button component
* style: update export button
* a11y: ExportAndShareButton
* add border
* quick style fix
* fix: flick issue on convo
* fix: DropDown opens when renaming
* chore: update radix-ui/react-dropdown-menu to latest
* small fix
* style: bookmarks update
* reorder export modal
* feat: imporved dropdowns
* style: a lot of changes; header, bookmarks, export, nav, convo, convoOptions
* fix: small style issues
* fix: button
* fix: bookmarks header menu
* fix: dropdown close glitch
* feat: Improve accessibility and keyboard navigation in ModelSpec component
* fix: Nav related type issues
* style: ConvoOptions theming and focus ring
---------
Co-authored-by: Danny Avila <danny@librechat.ai>
* refactor: scrollToEnd
* fix(validateConvoAccess): search conversation by ID for proper validation
* feat: Add unique index for conversationId and user in convoSchema
* refactor: Update font sizes 1 rem -> font-size-base in style.css
* fix: Assistants map type issues
* refactor: Remove obsolete scripts
* fix: Update DropdownNoState component to handle both string and OptionType values
* refactor: Remove config/loader.js file
* fix: remove crypto.randomBytes(); refactor: Create reusable function for generating token and hash
* fix: Update health endpoint URL
* refactor: use Constants for saved tag default value, do not place Saved as first always
* refactor: check trimmed currentText before appending parsedText in useSubmitMessage
* refactor: move `scrollToEnd()` to `createdHandler` and increase delay before execution slightly
* chore: Add back TypeScript linting rules for unnecessary conditions and strict boolean expressions
* chore: Update librechat-data-provider package.json version to 0.7.4.0
* feat: gpt-4o-2024-08-06 pricing for tx
* feat: add AWS models to tokenValues in tx.js for pricing transactions
* feat: Update tokenValues in tx.js for AWS models pricing
* refactor: add bedrock prefix values as well (temporary until we update value keys which includes context)
* refactor: optimize tag methods, remove rebuild
* refactor(tags): add lean db operations, fix updateTagsForConversation, remove rebuild button, only send convoId once
* refactor: Update BookmarkMenu to use Constants.NEW_CONVO constant for comparison
* style: Update BookmarkMenu styles and constants, use theming
* refactor: move tags query from package to client workspace
* refactor: optimize ConversationTag document creation and update logic
* style: Update BookmarkMenuItems to use theming
* refactor: JSDocs + try/catch for conversation tags API routes
* refactor: Update BookmarkNav theming classes and new data provider location
* fix: statefulness of conversation bookmarks
- move non-mutation hook to hooks/Conversation
- remove use of deprecated global convo
- update convo infinite data as well as current convo state upon successful tag add
* refactor: Update BookmarkMenu styles and constants, use theming
* refactor: Add lean option to ConversationTag deletion query
* fix(BookmarkTable): position order rendering esp. when new tag is created
* refactor: Update useBookmarkSucess to useBookmarkSuccess for consistency
* refactor: Update ConversationTag creation logic to increment count only if addToConversation is true
* style: theming
* 🛂 feat: Added Security for Conversation Access
* refactor: Update concurrentLimiter and convoAccess middleware to use isEnabled function for Redis check
* refactor: handle access check even if cache is not available (edge case)
* fix: correctly handle openai assistants annotations
* fix: Handle adjacent identical citations only for our specific format
* fix: correctly handle adjacent identical citations
* refactor: make regex handling more robust
* fix: skip annotation condition, make regex handling more robust
* refactor: Handle FILE_PATH and FILE_CITATION annotation types in processMessages.spec.js
* ci: unit tests for real file path type annotations
* fix: get user-provided API key using environment variable names as keys
* feat: Add error handling for missing API key and search engine ID
* feat: Add GoogleSearch and TavilySearchResults specs for environment variables handling
---------
Co-authored-by: Dongwoo Jeong <dongwoo.jeong@lge.com>
* fix: regex issue extracting text with images in markdown
* fix: update addImages function to include only the first observed image path in the response message
* ci: tests for addImages function: correct image extraction
* fix(GoogleClient): linting
---------
Co-authored-by: Dongwoo Jeong <dongwoo.jeong@lge.com>
Co-authored-by: Dongwoo Jeong <dongwoo@duck.com>
* chore: playwright setup update
* refactor: update ChatRoute component with accessible loading spinner with live region
* chore(Message): typing
* ci: first pass, a11y testing
* refactor: update lang attribute in index.html to "en-US"
* ci: jsx-a11y dev eslint plugin
* ci: jsx plugin
* fix: Exclude 'vite.config.ts' from TypeScript compilation for testing
* fix(a11y): Remove tabIndex from non-interactive element in MessagesView component
* fix(a11y):
- Visible, non-interactive elements with click handlers must have at least one keyboard listener.eslintjsx-a11y/click-events-have-key-events
- Avoid non-native interactive elements. If using native HTML is not possible, add an appropriate role and support for tabbing, mouse, keyboard, and touch inputs to an interactive content element.eslintjsx-a11y/no-static-element-interactions
chore: remove unused bookmarks panel
- fix some "Unexpected nullable boolean value in conditional" warnings
* fix(NewChat): a11y, nested button issue, add aria-label, remove implicit role
* fix(a11y):
- partially address #3515 with `main` landmark
other:
- eslint@typescript-eslint/strict-boolean-expressions
* chore(MenuButton): Use button element instead of div for accessibility
* chore: Update TitleButton to use button element for accessibility
* chore: Update TitleButton to use button element for accessibility
* refactor(ChatMenuItem): Improve focus accessibility and code readability
* chore(MenuButton): Update aria-label to dynamically include primaryText
* fix(a11y): SearchBar
- If a form control does not have a properly associated text label, the function or purpose of that form control may not be presented to screen reader users. Visible form labels also provide visible descriptions and larger clickable targets for form controls which placeholders do not.
* chore: remove duplicate SearchBar twcss
* fix(a11y):
- The edit and copy buttons that are visually hidden are exposed to Assistive technology and are announced to screen reader users.
* fix(a11y): visible focus outline
* fix(a11y): The button to select the LLM Model has the aria-haspopup and aria- expanded attributes which makes its role ambuguous and unclear. It functions like a combobox but doesn't fully support that interaction and also fucntions like a dialog but doesn't completely support that interaction either.
* fix(a11y): fix visible focus outline
* fix(a11y): Scroll to bottom button missing accessible name #3474
* fix(a11y): The page lacks any heading structure. There should be at least one H1 and other headings to help users understand the orgainzation of the page and the contents.
Note: h1 won't be correct here so made it h2
* fix(a11y): LLM controls aria-labels
* fix(a11y): There is no visible focus outline to the 'send message' button
* fix(a11y): fix visible focus outline for Fork button
* refactor(MessageRender): add focus ring to message cards, consolidate complex conditions, add logger for setting latest message, add tabindex for card
* fix: focus border color and fix set latest message card condition
* fix(a11y): Adequate contrast for MessageAudio buttton
* feat: Add GitHub Actions workflow for accessibility linting
* chore: Update GitHub Actions workflow for accessibility linting to include client/src/** path
* fix(Nav): navmask and accessibility
* fix: Update Nav component to handle potential undefined type in SearchContext
* fix(a11y): add focus visibility to attach files button #3475
* fix(a11y): discernible text for NewChat button
* fix(a11y): accessible landmark names, all page content in landmarks, ensures landmarks are unique #3514#3515
* fix(Prompts): update isChatRoute prop to be required in List component
* fix(a11y): buttons must have discernible text
- Introduce getSafetySettings method for centralized safety settings
- Apply safety settings uniformly in sendCompletion for all models, including Gemini
- Ensure consistent safety settings application in getCompletion
- Update .env.example to clarify safety settings usage for both APIs
* feat: add tags property in Conversation model
* feat: add ConversationTag model
* feat: add the tags parameter to getConvosByPage
* feat: add API route to ConversationTag
* feat: add types of ConversationTag
* feat: add data access functions for conversation tags
* feat: add Bookmark table component
* feat: Add an action to bookmark
* feat: add Bookmark nav component
* fix: failed test
* refactor: made 'Saved' tag a constant
* feat: add new bookmark to current conversation
* chore: Add comment
* fix: delete tag from conversations when it's deleted
* fix: Update the query cache when the tag title is changed.
* chore: fix typo
* refactor: add description of rebuilding bookmarks
* chore: remove unused variables
* fix: position when adding a new bookmark
* refactor: add comment, rename a function
* refactor: add a unique constraint in ConversationTag
* chore: add localizations
* Allow LDAP login via username
This patch adds the option to login via username instead of using an
email address since the latter may not be unique or may change.
For example, our organization has two main domains and users have a log
and a short form of their mail address. This makes it hard for users to
identify what their primary email address is and causes a lot of
confusion. Using their username instead makes it much easier.
Using a username will also make it easier in the future to not need a
separate bind user to get user attributes. So, this is also a bit of
prep work for that.
* Update config.js
* feat: Enable LDAP login via username
This commit enables the option to login via username instead of using an email address for LDAP authentication. This change is necessary because email addresses may not be unique or may change, causing confusion for users. By using usernames, it becomes easier for users to identify their primary email address. Additionally, this change prepares for future improvements by eliminating the need for a separate bind user to retrieve user attributes.
Co-authored-by: Danny Avila <danny@librechat.ai>
* chore: jsdocs
* chore: import order
* ci: add ldap config tests
---------
Co-authored-by: Lars Kiesow <lkiesow@uos.de>
* chore: bump data-provider
* feat: Add anthropicSettings to endpointSettings
The commit adds the `anthropicSettings` object to the `endpointSettings` in the `schemas.ts` file. This allows for the configuration of settings specific to the `anthropic` model endpoint.
* chore: adjust maxoutputtokens localization
* feat: Update AnthropicClient to use anthropicSettings for default model options and increased output beta header
* ci: new anthropic tests
- Add unique index for messageId and user in messageSchema
- use `updateMessage` for updating the plugins message?
- add better logging for updateMessage
- prevents dupe_key or getKeyIndex error
* feat: `gpt-4o-mini`
* feat: retrival
* fix: Update order of model token values for 'gpt-4o' and 'gpt-4o-mini'
* fix: Update order of model token values for 'gpt-4o' and 'gpt-4o-mini'
* fix: Update order of model token values for 'gpt-4o' and 'gpt-4o-mini'
* fix: add jsdoc
* fix: Update order of model token values for 'gpt-4o' and 'gpt-4o-mini'
---------
Co-authored-by: Danny Avila <danny@librechat.ai>
* fix: edge case with debugTraverse function in parsers.js
* chore: Update error message in audio stream processing
* refactor: Add ONE_MINUTE and THIRTY_SECONDS options to Time enum
* fix: logging
* fix: do not throw errors for invalid convo id, simply return undefined
* feat: Add error handling and logging to message route definitions
* feat: Refactor Message.js to improve code organization and readability
* fix: Return undefined instead of throwing error for invalid conversation ID in Message.spec.js
includedTools expects the pluginKey/tool.name for filtering included tools, which is incompatible with the previous implementation in loadAndFormatTools used for loading assistant tools.
This change uses both filename and tool.name for filtering tools in ToolService.loadAndFormatTools in order to make it compatible with the old implementation and the behaviour of plugin filtering.
Co-authored-by: Benedikt Nordhoff <benedikt.nordhoff@codecentric.de>
* fix(Speech): speechTab settings update
* fix: get speech config; refactor: moved everything to types and removed file types; feature: not overriding variables selected by user
