{"entries":[{"lastCommit":{"author":{"email":"danny@librechat.ai","name":"Danny Avila","when":"2025-11-10T19:05:30.000000-05:00"},"committer":{"email":"noreply@github.com","name":"GitHub","when":"2025-11-10T19:05:30.000000-05:00"},"hash":"2524d3336264084caa9cc5e909186e10f00adbf5","message":"📂 refactor: Cleanup File Filtering Logic, Improve Validation (#10414)\n\n* feat: add filterFilesByEndpointConfig to filter disabled file processing by provider\n\n* chore: explicit define of endpointFileConfig for better debugging\n\n* refactor: move `normalizeEndpointName` to data-provider as used app-wide\n\n* chore: remove overrideEndpoint from useFileHandling\n\n* refactor: improve endpoint file config selection\n\n* refactor: update filterFilesByEndpointConfig to accept structured parameters and improve endpoint file config handling\n\n* refactor: replace defaultFileConfig with getEndpointFileConfig for improved file configuration handling across components\n\n* test: add comprehensive unit tests for getEndpointFileConfig to validate endpoint configuration handling\n\n* refactor: streamline agent endpoint assignment and improve file filtering logic\n\n* feat: add error handling for disabled file uploads in endpoint configuration\n\n* refactor: update encodeAndFormat functions to accept structured parameters for provider and endpoint\n\n* refactor: streamline requestFiles handling in initializeAgent function\n\n* fix: getEndpointFileConfig partial config merging scenarios\n\n* refactor: enhance mergeWithDefault function to support document-supported providers with comprehensive MIME types\n\n* refactor: user-configured default file config in getEndpointFileConfig\n\n* fix: prevent file handling when endpoint is disabled and file is dragged to chat\n\n* refactor: move `getEndpointField` to `data-provider` and update usage across components and hooks\n\n* fix: prioritize endpointType based on agent.endpoint in file filtering logic\n\n* fix: prioritize agent.endpoint in file filtering logic and remove unnecessary endpointType defaulting","tree":"b0381b1e7e208a563e68d1b803d23199a02364b7"},"mode":"0040000","name":"agents","size":0},{"lastCommit":{"author":{"email":"danny@librechat.ai","name":"Danny Avila","when":"2025-11-05T17:15:17.000000-05:00"},"committer":{"email":"noreply@github.com","name":"GitHub","when":"2025-11-05T17:15:17.000000-05:00"},"hash":"8a4a5a47903d354daab75bb26ea72b229b5a0571","message":"🤖 feat: Agent Handoffs (Routing) (#10176)\n\n* feat: Add support for agent handoffs with edges in agent forms and schemas\n\nchore: Mark `agent_ids` field as deprecated in favor of edges across various schemas and types\n\nchore: Update dependencies for @langchain/core and @librechat/agents to latest versions\n\nchore: Update peer dependency for @librechat/agents to version 3.0.0-rc2 in package.json\n\nchore: Update @librechat/agents dependency to version 3.0.0-rc3 in package.json and package-lock.json\n\nfeat: first pass, multi-agent handoffs\n\nfix: update output type to ToolMessage in memory handling functions\n\nfix: improve type checking for graphConfig in createRun function\n\nrefactor: remove unused content filtering logic in AgentClient\n\nchore: update @librechat/agents dependency to version 3.0.0-rc4 in package.json and package-lock.json\n\nfix: update @langchain/core peer dependency version to ^0.3.72 in package.json and package-lock.json\n\nfix: update @librechat/agents dependency to version 3.0.0-rc6 in package.json and package-lock.json; refactor stream rate handling in various endpoints\n\nfeat: Agent handoff UI\n\nchore: update @librechat/agents dependency to version 3.0.0-rc8 in package.json and package-lock.json\n\nfix: improve hasInfo condition and adjust UI element classes in AgentHandoff component\n\nrefactor: remove current fixed agent display from AgentHandoffs component due to redundancy\n\nfeat: enhance AgentHandoffs UI with localized beta label and improved layout\n\nchore: update @librechat/agents dependency to version 3.0.0-rc10 in package.json and package-lock.json\n\nfeat: add `createSequentialChainEdges` function to add back agent chaining via multi-agents\n\nfeat: update `createSequentialChainEdges` call to only provide conversation context between agents\n\nfeat: deprecate Agent Chain functionality and update related methods for improved clarity\n\n* chore: update @librechat/agents dependency to version 3.0.0-rc11 in package.json and package-lock.json\n\n* refactor: remove unused addCacheControl function and related imports and import from @librechat/agents\n\n* chore: remove unused i18n keys\n\n* refactor: remove unused format export from index.ts\n\n* chore: update @librechat/agents to v3.0.0-rc13\n\n* chore: remove BEDROCK_LEGACY provider from Providers enum\n\n* chore: update @librechat/agents to version 3.0.2 in package.json","tree":"0800d4a37dce6acb142d710726e7936c0b7c022a"},"mode":"0040000","name":"anthropic","size":0},{"lastCommit":{"author":{"email":"danny@librechat.ai","name":"Danny Avila","when":"2025-11-28T11:14:13.000000-05:00"},"committer":{"email":"noreply@github.com","name":"GitHub","when":"2025-11-28T11:14:13.000000-05:00"},"hash":"801c95a829a29e5182a4fac3f8ff2f7725ca38ae","message":"🦙 fix: Ollama Provider Handling (#10711)\n\n* 🔧 fix: Correct URL Construction in fetchModels Function\n\nUpdated the URL construction in the fetchModels function to ensure proper formatting by removing trailing slashes from the base URL. This change prevents potential issues with API endpoint calls.\n\n* 🔧 fix: Remove OLLAMA from Known Custom Providers\n\nUpdated the isKnownCustomProvider function and providerConfigMap to exclude OLLAMA as a known custom provider, streamlining the provider checks and configurations.\n\n* 🔧 test: Enhance fetchModels Tests for URL Construction\n\nAdded new test cases to validate the URL construction in the fetchModels function, ensuring it handles trailing slashes correctly and appends query parameters as expected. This improves the robustness of the API endpoint calls.\n\n* chore: remove ollama provider-specific handling\n\n* chore: Refactor imports to use isUserProvided from @librechat/api","tree":"7fda77360e220cdf913b4d37fd040c22a8b64533"},"mode":"0040000","name":"assistants","size":0},{"lastCommit":{"author":{"email":"danny@librechat.ai","name":"Danny Avila","when":"2025-08-26T12:10:18.000000-04:00"},"committer":{"email":"noreply@github.com","name":"GitHub","when":"2025-08-26T12:10:18.000000-04:00"},"hash":"9a210971f530de1650d9afaf65d5dab6f3e97f98","message":"🛜 refactor: Streamline App Config Usage (#9234)\n\n* WIP: app.locals refactoring\n\nWIP: appConfig\n\nfix: update memory configuration retrieval to use getAppConfig based on user role\n\nfix: update comment for AppConfig interface to clarify purpose\n\n🏷️ refactor: Update tests to use getAppConfig for endpoint configurations\n\nci: Update AppService tests to initialize app config instead of app.locals\n\nci: Integrate getAppConfig into remaining tests\n\nrefactor: Update multer storage destination to use promise-based getAppConfig and improve error handling in tests\n\nrefactor: Rename initializeAppConfig to setAppConfig and update related tests\n\nci: Mock getAppConfig in various tests to provide default configurations\n\nrefactor: Update convertMCPToolsToPlugins to use mcpManager for server configuration and adjust related tests\n\nchore: rename `Config/getAppConfig` -> `Config/app`\n\nfix: streamline OpenAI image tools configuration by removing direct appConfig dependency and using function parameters\n\nchore: correct parameter documentation for imageOutputType in ToolService.js\n\nrefactor: remove `getCustomConfig` dependency in config route\n\nrefactor: update domain validation to use appConfig for allowed domains\n\nrefactor: use appConfig registration property\n\nchore: remove app parameter from AppService invocation\n\nrefactor: update AppConfig interface to correct registration and turnstile configurations\n\nrefactor: remove getCustomConfig dependency and use getAppConfig in PluginController, multer, and MCP services\n\nrefactor: replace getCustomConfig with getAppConfig in STTService, TTSService, and related files\n\nrefactor: replace getCustomConfig with getAppConfig in Conversation and Message models, update tempChatRetention functions to use AppConfig type\n\nrefactor: update getAppConfig calls in Conversation and Message models to include user role for temporary chat expiration\n\nci: update related tests\n\nrefactor: update getAppConfig call in getCustomConfigSpeech to include user role\n\nfix: update appConfig usage to access allowedDomains from actions instead of registration\n\nrefactor: enhance AppConfig to include fileStrategies and update related file strategy logic\n\nrefactor: update imports to use normalizeEndpointName from @librechat/api and remove redundant definitions\n\nchore: remove deprecated unused RunManager\n\nrefactor: get balance config primarily from appConfig\n\nrefactor: remove customConfig dependency for appConfig and streamline loadConfigModels logic\n\nrefactor: remove getCustomConfig usage and use app config in file citations\n\nrefactor: consolidate endpoint loading logic into loadEndpoints function\n\nrefactor: update appConfig access to use endpoints structure across various services\n\nrefactor: implement custom endpoints configuration and streamline endpoint loading logic\n\nrefactor: update getAppConfig call to include user role parameter\n\nrefactor: streamline endpoint configuration and enhance appConfig usage across services\n\nrefactor: replace getMCPAuthMap with getUserMCPAuthMap and remove unused getCustomConfig file\n\nrefactor: add type annotation for loadedEndpoints in loadEndpoints function\n\nrefactor: move /services/Files/images/parse to TS API\n\nchore: add missing FILE_CITATIONS permission to IRole interface\n\nrefactor: restructure toolkits to TS API\n\nrefactor: separate manifest logic into its own module\n\nrefactor: consolidate tool loading logic into a new tools module for startup logic\n\nrefactor: move interface config logic to TS API\n\nrefactor: migrate checkEmailConfig to TypeScript and update imports\n\nrefactor: add FunctionTool interface and availableTools to AppConfig\n\nrefactor: decouple caching and DB operations from AppService, make part of consolidated `getAppConfig`\n\nWIP: fix tests\n\n* fix: rebase conflicts\n\n* refactor: remove app.locals references\n\n* refactor: replace getBalanceConfig with getAppConfig in various strategies and middleware\n\n* refactor: replace appConfig?.balance with getBalanceConfig in various controllers and clients\n\n* test: add balance configuration to titleConvo method in AgentClient tests\n\n* chore: remove unused `openai-chat-tokens` package\n\n* chore: remove unused imports in initializeMCPs.js\n\n* refactor: update balance configuration to use getAppConfig instead of getBalanceConfig\n\n* refactor: integrate configMiddleware for centralized configuration handling\n\n* refactor: optimize email domain validation by removing unnecessary async calls\n\n* refactor: simplify multer storage configuration by removing async calls\n\n* refactor: reorder imports for better readability in user.js\n\n* refactor: replace getAppConfig calls with req.config for improved performance\n\n* chore: replace getAppConfig calls with req.config in tests for centralized configuration handling\n\n* chore: remove unused override config\n\n* refactor: add configMiddleware to endpoint route and replace getAppConfig with req.config\n\n* chore: remove customConfig parameter from TTSService constructor\n\n* refactor: pass appConfig from request to processFileCitations for improved configuration handling\n\n* refactor: remove configMiddleware from endpoint route and retrieve appConfig directly in getEndpointsConfig if not in `req.config`\n\n* test: add mockAppConfig to processFileCitations tests for improved configuration handling\n\n* fix: pass req.config to hasCustomUserVars and call without await after synchronous refactor\n\n* fix: type safety in useExportConversation\n\n* refactor: retrieve appConfig using getAppConfig in PluginController and remove configMiddleware from plugins route, to avoid always retrieving when plugins are cached\n\n* chore: change `MongoUser` typedef to `IUser`\n\n* fix: Add `user` and `config` fields to ServerRequest and update JSDoc type annotations from Express.Request to ServerRequest\n\n* fix: remove unused setAppConfig mock from Server configuration tests","tree":"704bbf88f3f5e635e0b2835b27ad3817485a069f"},"mode":"0040000","name":"azureAssistants","size":0},{"lastCommit":{"author":{"email":"danny@librechat.ai","name":"Danny Avila","when":"2025-11-21T16:42:28.000000-05:00"},"committer":{"email":"noreply@github.com","name":"GitHub","when":"2025-11-21T16:42:28.000000-05:00"},"hash":"35319c135459be9580cee97ef7d72e225526592a","message":"🔧 fix: Remove Bedrock Config Transform introduced in #9931 (#10628)\n\n* fix: Header and Environment Variable Handling Bug from #9931\n\n* refactor: Remove warning log for missing tokens in extractOpenIDTokenInfo function\n\n* feat: Enhance resolveNestedObject function for improved placeholder processing\n\n- Added a new function `resolveNestedObject` to recursively process nested objects, replacing placeholders in string values while preserving the original structure.\n- Updated `createTestUser` to use `IUser` type and modified user ID generation.\n- Added comprehensive unit tests for `resolveNestedObject` to cover various scenarios, including nested structures, arrays, and custom user variables.\n- Improved type handling in `processMCPEnv` to ensure correct processing of mixed numeric and placeholder values.\n\n* refactor: Remove unnecessary manipulation of Bedrock options introduced in #9931\n\n- Eliminated the resolveHeaders function call from the getOptions method in options.js, as it was no longer necessary for processing additional model request fields.\n- This change simplifies the code and improves maintainability.","tree":"f7cb7d0cb8c23575f670b6ca293d17a79856b094"},"mode":"0040000","name":"bedrock","size":0},{"lastCommit":{"author":{"email":"levy@andes.is","name":"Jón Levy","when":"2025-11-21T14:44:52.000000Z"},"committer":{"email":"danny@librechat.ai","name":"Danny Avila","when":"2025-11-21T09:51:11.000000-05:00"},"hash":"ef3bf0a9321eb8b0a3c1cab7b8b2cec4794ab346","message":"🆔 feat: Add OpenID Connect Federated Provider Token Support (#9931)\n\n* feat: Add OpenID Connect federated provider token support\n\nImplements support for passing federated provider tokens (Cognito, Azure AD, Auth0)\nas variables in LibreChat's librechat.yaml configuration for both custom endpoints\nand MCP servers.\n\nFeatures:\n- New LIBRECHAT_OPENID_* template variables for federated provider tokens\n- JWT claims parsing from ID tokens without verification (for claim extraction)\n- Token validation with expiration checking\n- Support for multiple token storage locations (federatedTokens, openidTokens)\n- Integration with existing template variable system\n- Comprehensive test suite with Cognito-specific scenarios\n- Provider-agnostic design supporting Cognito, Azure AD, Auth0, etc.\n\nSecurity:\n- Server-side only token processing\n- Automatic token expiration validation\n- Graceful fallbacks for missing/invalid tokens\n- No client-side token exposure\n\n🤖 Generated with [Claude Code](https://claude.com/claude-code)\n\nCo-Authored-By: Claude \n\n* fix: Add federated token propagation to OIDC authentication strategies\n\nAdds federatedTokens object to user during authentication to enable\nfederated provider token template variables in LibreChat configuration.\n\nChanges:\n- OpenID JWT Strategy: Extract raw JWT from Authorization header and\n attach as federatedTokens.access_token to enable {{LIBRECHAT_OPENID_TOKEN}}\n placeholder resolution\n- OpenID Strategy: Attach tokenset tokens as federatedTokens object to\n standardize token access across both authentication strategies\n\nThis enables proper token propagation for custom endpoints and MCP\nservers that require federated provider tokens for authorization.\n\nResolves missing token issue reported by @ramden in PR #9931\n\n🤖 Generated with [Claude Code](https://claude.com/claude-code)\n\nCo-Authored-By: Denis Ramic \nCo-Authored-By: Claude \n\n* test: Add federatedTokens validation tests for OIDC strategies\n\nAdds comprehensive test coverage for the federated token propagation\nfeature implemented in the authentication strategies.\n\nTests added:\n- Verify federatedTokens object is attached to user with correct structure\n (access_token, refresh_token, expires_at)\n- Verify both tokenset and federatedTokens are present in user object\n- Ensure tokens from OIDC provider are correctly propagated\n\nAlso fixes existing test suite by adding missing mocks:\n- isEmailDomainAllowed function mock\n- findOpenIDUser function mock\n\nThese tests validate the fix from commit 5874ba29f that enables\n{{LIBRECHAT_OPENID_TOKEN}} template variable functionality.\n\n🤖 Generated with [Claude Code](https://claude.com/claude-code)\n\nCo-Authored-By: Claude \n\n* docs: Remove implementation documentation file\n\nThe PR description already contains all necessary implementation details.\nThis documentation file is redundant and was requested to be removed.\n\n* fix: skip s256 check\n\n* fix(openid): handle missing refresh token in Cognito token refresh response\n\nWhen OPENID_REUSE_TOKENS=true, the token refresh flow was failing because\nCognito (and most OAuth providers) don't return a new refresh token in the\nrefresh grant response - they only return new access and ID tokens.\n\nChanges:\n- Modified setOpenIDAuthTokens() to accept optional existingRefreshToken parameter\n- Updated validation to only require access_token (refresh_token now optional)\n- Added logic to reuse existing refresh token when not provided in tokenset\n- Updated refreshController to pass original refresh token as fallback\n- Added comments explaining standard OAuth 2.0 refresh token behavior\n\nThis fixes the \"Token is not present. User is not authenticated.\" error that\noccurred during silent token refresh with Cognito as the OpenID provider.\n\nFixes: Authentication loop with OPENID_REUSE_TOKENS=true and AWS Cognito\n\n* fix(openid): extract refresh token from cookies for template variable replacement\n\nWhen OPENID_REUSE_TOKENS=true, the openIdJwtStrategy populates user.federatedTokens\nto enable template variable replacement (e.g., {{LIBRECHAT_OPENID_ACCESS_TOKEN}}).\n\nHowever, the refresh_token field was incorrectly sourced from payload.refresh_token,\nwhich is always undefined because:\n1. JWTs don't contain refresh tokens in their payload\n2. The JWT itself IS the access token\n3. Refresh tokens are separate opaque tokens stored in HTTP-only cookies\n\nThis caused extractOpenIDTokenInfo() to receive incomplete federatedTokens,\nresulting in template variables remaining unreplaced in headers.\n\n**Root Cause:**\n- Line 90: `refresh_token: payload.refresh_token` (always undefined)\n- JWTs only contain access token data in their claims\n- Refresh tokens are separate, stored securely in cookies\n\n**Solution:**\n- Import `cookie` module to parse cookies from request\n- Extract refresh token from `refreshToken` cookie\n- Populate federatedTokens with both access token (JWT) and refresh token (from cookie)\n\n**Impact:**\n- Template variables like {{LIBRECHAT_OPENID_ACCESS_TOKEN}} now work correctly\n- Headers in librechat.yaml are properly replaced with actual tokens\n- MCP server authentication with federated tokens now functional\n\n**Technical Details:**\n- passReqToCallback=true in JWT strategy provides req object access\n- Refresh token extracted via cookies.parse(req.headers.cookie).refreshToken\n- Falls back gracefully if cookie header or refreshToken is missing\n\n🤖 Generated with [Claude Code](https://claude.com/claude-code)\n\nCo-Authored-By: Claude \n\n* fix: re-resolve headers on each request to pick up fresh federatedTokens\n\n- OpenAIClient now re-resolves headers in chatCompletion() before each API call\n- This ensures template variables like {{LIBRECHAT_OPENID_TOKEN}} are replaced\n with actual token values from req.user.federatedTokens\n- initialize.js now stores original template headers instead of pre-resolved ones\n- Fixes template variable replacement when OPENID_REUSE_TOKENS=true\n\nThe issue was that headers were only resolved once during client initialization,\nbefore openIdJwtStrategy had populated user.federatedTokens. Now headers are\nre-resolved on every request with the current user's fresh tokens.\n\n* debug: add logging to track header resolution in OpenAIClient\n\n* debug: log tokenset structure after refresh to diagnose missing access_token\n\n* fix: set federatedTokens on user object after OAuth refresh\n\n- After successful OAuth token refresh, the user object was not being\n updated with federatedTokens\n- This caused template variable resolution to fail on subsequent requests\n- Now sets user.federatedTokens with access_token, id_token, refresh_token\n and expires_at from the refreshed tokenset\n- Fixes template variables like {{LIBRECHAT_OPENID_TOKEN}} not being\n replaced after token refresh\n- Related to PR #9931 (OpenID federated token support)\n\n* fix(openid): pass user object through agent chain for template variable resolution\n\nRoot cause: buildAgentContext in agents/run.ts called resolveHeaders without\nthe user parameter, preventing OpenID federated token template variables from\nbeing resolved in agent runtime parameters.\n\nChanges:\n- packages/api/src/agents/run.ts: Add user parameter to createRun signature\n- packages/api/src/agents/run.ts: Pass user to resolveHeaders in buildAgentContext\n- api/server/controllers/agents/client.js: Pass user when calling createRun\n- api/server/services/Endpoints/bedrock/options.js: Add resolveHeaders call with debug logging\n- api/server/services/Endpoints/custom/initialize.js: Add debug logging\n- packages/api/src/utils/env.ts: Add comprehensive debug logging and stack traces\n- packages/api/src/utils/oidc.ts: Fix eslint errors (unused type, explicit any)\n\nThis ensures template variables like {{LIBRECHAT_OPENID_TOKEN}} and\n{{LIBRECHAT_USER_OPENIDID}} are properly resolved in both custom endpoint\nheaders and Bedrock AgentCore runtime parameters.\n\n🤖 Generated with [Claude Code](https://claude.com/claude-code)\n\nCo-Authored-By: Claude \n\n* refactor: remove debug logging from OpenID token template feature\n\nRemoved excessive debug logging that was added during development to make\nthe PR more suitable for upstream review:\n\n- Removed 7 debug statements from OpenAIClient.js\n- Removed all console.log statements from packages/api/src/utils/env.ts\n- Removed debug logging from bedrock/options.js\n- Removed debug logging from custom/initialize.js\n- Removed debug statement from AuthController.js\n\nThis reduces the changeset by ~50 lines while maintaining full functionality\nof the OpenID federated token template variable feature.\n\n🤖 Generated with [Claude Code](https://claude.com/claude-code)\n\nCo-Authored-By: Claude \n\n* test(openid): add comprehensive unit tests for template variable substitution\n\n- Add 34 unit tests for OIDC token utilities (oidc.spec.ts)\n- Test coverage for token extraction, validation, and placeholder processing\n- Integration tests for full OpenID token flow\n- All tests pass with comprehensive edge case coverage\n\n🤖 Generated with Claude Code\nCo-Authored-By: Claude \n\n* test: fix OpenID federated tokens test failures\n\n- Add serverMetadata() mock to openid-client mock configuration\n * Fixes TypeError in openIdJwtStrategy.js where serverMetadata() was being called\n * Mock now returns jwks_uri and end_session_endpoint as expected by the code\n\n- Update outdated initialize.spec.js test\n * Remove test expecting resolveHeaders call during initialization\n * Header resolution was refactored to be deferred until LLM request time\n * Update test to verify options are returned correctly with useLegacyContent flag\n\nFixes #9931 CI failures for backend unit tests\n\n🤖 Generated with [Claude Code](https://claude.com/claude-code)\n\nCo-Authored-By: Claude \n\n* chore: fix package-lock.json conflict\n\n* chore: sync package-log with upstream\n\n* chore: cleanup\n\n* fix: use createSafeUser\n\n* fix: fix createSafeUser signature\n\n* chore: remove comments\n\n* chore: purge comments\n\n* fix: update Jest testPathPattern to testPathPatterns for Jest 30+ compatibility\n\n---------\n\nCo-authored-by: Claude \nCo-authored-by: Denis Ramic \nCo-authored-by: kristjanaapro \n\nchore: import order and add back JSDoc for OpenID JWT callback\n","tree":"9eb3a8c74c976e9215748d12376aaa3842186a35"},"mode":"0040000","name":"custom","size":0},{"lastCommit":{"author":{"email":"danny@librechat.ai","name":"Danny Avila","when":"2025-08-27T03:44:39.000000-04:00"},"committer":{"email":"noreply@github.com","name":"GitHub","when":"2025-08-27T03:44:39.000000-04:00"},"hash":"48f6f8f2f802d99b910bb67ca8c264254cf9ce62","message":"📎 feat: Upload as Text Support for Plaintext, STT, RAG, and Token Limits (#8868)\n\n* 🪶 feat: Add Support for Uploading Plaintext Files\n\nfeat: delineate between OCR and text handling in fileConfig field of config file\n\n- also adds support for passing in mimetypes as just plain file extensions\n\nfeat: add showLabel bool to support future synthetic component DynamicDropdownInput\n\nfeat: add new combination dropdown-input component in params panel to support file type token limits\n\nrefactor: move hovercard to side to align with other hovercards\n\nchore: clean up autogenerated comments\n\nfeat: add delineation to file upload path between text and ocr configured filetypes\n\nfeat: add token limit checks during file upload\n\nrefactor: move textParsing out of ocrEnabled logic\n\nrefactor: clean up types for filetype config\n\nrefactor: finish decoupling DynamicDropdownInput from fileTokenLimits\n\nfix: move image token cost function into file to fix circular dependency causing unittest to fail and remove unused var for linter\n\nchore: remove out of scope code following review\n\nrefactor: make fileTokenLimit conform to existing styles\n\nchore: remove unused localization string\n\nchore: undo changes to DynamicInput and other strays\n\nfeat: add fileTokenLimit to all provider config panels\n\nfix: move textParsing back into ocr tool_resource block for now so that it doesn't interfere with other upload types\n\n* 📤 feat: Add RAG API Endpoint Support for Text Parsing (#8849)\n\n* feat: implement RAG API integration for text parsing with fallback to native parsing\n\n* chore: remove TODO now that placeholder and fllback are implemented\n\n* ✈️ refactor: Migrate Text Parsing to TS (#8892)\n\n* refactor: move generateShortLivedToken to packages/api\n\n* refactor: move textParsing logic into packages/api\n\n* refactor: reduce nesting and dry code with createTextFile\n\n* fix: add proper source handling\n\n* fix: mock new parseText and parseTextNative functions in jest file\n\n* ci: add test coverage for textParser\n\n* 💬 feat: Add Audio File Support to Upload as Text (#8893)\n\n* feat: add STT support for Upload as Text\n\n* refactor: move processAudioFile to packages/api\n\n* refactor: move textParsing from utils to files\n\n* fix: remove audio/mp3 from unsupported mimetypes test since it is now supported\n\n* ✂️ feat: Configurable File Token Limits and Truncation (#8911)\n\n* feat: add configurable fileTokenLimit default value\n\n* fix: add stt to fileConfig merge logic\n\n* fix: add fileTokenLimit to mergeFileConfig logic so configurable value is actually respected from yaml\n\n* feat: add token limiting to parsed text files\n\n* fix: add extraction logic and update tests so fileTokenLimit isnt sent to LLM providers\n\n* fix: address comments\n\n* refactor: rename textTokenLimiter.ts to text.ts\n\n* chore: update form-data package to address CVE-2025-7783 and update package-lock\n\n* feat: use default supported mime types for ocr on frontend file validation\n\n* fix: should be using logger.debug not console.debug\n\n* fix: mock existsSync in text.spec.ts\n\n* fix: mock logger rather than every one of its function calls\n\n* fix: reorganize imports and streamline file upload processing logic\n\n* refactor: update createTextFile function to use destructured parameters and improve readability\n\n* chore: update file validation to use EToolResources for improved type safety\n\n* chore: update import path for types in audio processing module\n\n* fix: update file configuration access and replace console.debug with logger.debug for improved logging\n\n---------\n\nCo-authored-by: Dustin Healy \nCo-authored-by: Dustin Healy <54083382+dustinhealy@users.noreply.github.com>","tree":"c69694736c21f69ba222015656a476522ed86c7f"},"mode":"0040000","name":"google","size":0},{"lastCommit":{"author":{"email":"danny@librechat.ai","name":"Danny Avila","when":"2025-11-28T11:14:13.000000-05:00"},"committer":{"email":"noreply@github.com","name":"GitHub","when":"2025-11-28T11:14:13.000000-05:00"},"hash":"801c95a829a29e5182a4fac3f8ff2f7725ca38ae","message":"🦙 fix: Ollama Provider Handling (#10711)\n\n* 🔧 fix: Correct URL Construction in fetchModels Function\n\nUpdated the URL construction in the fetchModels function to ensure proper formatting by removing trailing slashes from the base URL. This change prevents potential issues with API endpoint calls.\n\n* 🔧 fix: Remove OLLAMA from Known Custom Providers\n\nUpdated the isKnownCustomProvider function and providerConfigMap to exclude OLLAMA as a known custom provider, streamlining the provider checks and configurations.\n\n* 🔧 test: Enhance fetchModels Tests for URL Construction\n\nAdded new test cases to validate the URL construction in the fetchModels function, ensuring it handles trailing slashes correctly and appends query parameters as expected. This improves the robustness of the API endpoint calls.\n\n* chore: remove ollama provider-specific handling\n\n* chore: Refactor imports to use isUserProvided from @librechat/api","tree":"7fda77360e220cdf913b4d37fd040c22a8b64533"},"mode":"0100644","name":"index.js","size":2769},{"lastCommit":{"author":{"email":"danny@librechat.ai","name":"Danny Avila","when":"2025-11-05T17:15:17.000000-05:00"},"committer":{"email":"noreply@github.com","name":"GitHub","when":"2025-11-05T17:15:17.000000-05:00"},"hash":"8a4a5a47903d354daab75bb26ea72b229b5a0571","message":"🤖 feat: Agent Handoffs (Routing) (#10176)\n\n* feat: Add support for agent handoffs with edges in agent forms and schemas\n\nchore: Mark `agent_ids` field as deprecated in favor of edges across various schemas and types\n\nchore: Update dependencies for @langchain/core and @librechat/agents to latest versions\n\nchore: Update peer dependency for @librechat/agents to version 3.0.0-rc2 in package.json\n\nchore: Update @librechat/agents dependency to version 3.0.0-rc3 in package.json and package-lock.json\n\nfeat: first pass, multi-agent handoffs\n\nfix: update output type to ToolMessage in memory handling functions\n\nfix: improve type checking for graphConfig in createRun function\n\nrefactor: remove unused content filtering logic in AgentClient\n\nchore: update @librechat/agents dependency to version 3.0.0-rc4 in package.json and package-lock.json\n\nfix: update @langchain/core peer dependency version to ^0.3.72 in package.json and package-lock.json\n\nfix: update @librechat/agents dependency to version 3.0.0-rc6 in package.json and package-lock.json; refactor stream rate handling in various endpoints\n\nfeat: Agent handoff UI\n\nchore: update @librechat/agents dependency to version 3.0.0-rc8 in package.json and package-lock.json\n\nfix: improve hasInfo condition and adjust UI element classes in AgentHandoff component\n\nrefactor: remove current fixed agent display from AgentHandoffs component due to redundancy\n\nfeat: enhance AgentHandoffs UI with localized beta label and improved layout\n\nchore: update @librechat/agents dependency to version 3.0.0-rc10 in package.json and package-lock.json\n\nfeat: add `createSequentialChainEdges` function to add back agent chaining via multi-agents\n\nfeat: update `createSequentialChainEdges` call to only provide conversation context between agents\n\nfeat: deprecate Agent Chain functionality and update related methods for improved clarity\n\n* chore: update @librechat/agents dependency to version 3.0.0-rc11 in package.json and package-lock.json\n\n* refactor: remove unused addCacheControl function and related imports and import from @librechat/agents\n\n* chore: remove unused i18n keys\n\n* refactor: remove unused format export from index.ts\n\n* chore: update @librechat/agents to v3.0.0-rc13\n\n* chore: remove BEDROCK_LEGACY provider from Providers enum\n\n* chore: update @librechat/agents to version 3.0.2 in package.json","tree":"0800d4a37dce6acb142d710726e7936c0b7c022a"},"mode":"0040000","name":"openAI","size":0}],"lastCommit":{"author":{"email":"danny@librechat.ai","name":"Danny Avila","when":"2025-11-28T11:14:13.000000-05:00"},"committer":{"email":"noreply@github.com","name":"GitHub","when":"2025-11-28T11:14:13.000000-05:00"},"hash":"801c95a829a29e5182a4fac3f8ff2f7725ca38ae","message":"🦙 fix: Ollama Provider Handling (#10711)\n\n* 🔧 fix: Correct URL Construction in fetchModels Function\n\nUpdated the URL construction in the fetchModels function to ensure proper formatting by removing trailing slashes from the base URL. This change prevents potential issues with API endpoint calls.\n\n* 🔧 fix: Remove OLLAMA from Known Custom Providers\n\nUpdated the isKnownCustomProvider function and providerConfigMap to exclude OLLAMA as a known custom provider, streamlining the provider checks and configurations.\n\n* 🔧 test: Enhance fetchModels Tests for URL Construction\n\nAdded new test cases to validate the URL construction in the fetchModels function, ensuring it handles trailing slashes correctly and appends query parameters as expected. This improves the robustness of the API endpoint calls.\n\n* chore: remove ollama provider-specific handling\n\n* chore: Refactor imports to use isUserProvided from @librechat/api","tree":"7fda77360e220cdf913b4d37fd040c22a8b64533"}}