1 Release published by 1 person

• v1.10.10-lts

  published Aug 21, 2025

36 Pull requests merged by 3 people

• style: MCP config dialog style

  #3911 merged Aug 21, 2025

• fix: Form collection file upload cannot delete file

  #3910 merged Aug 21, 2025

• feat: Vllm reranker model bge reranker v2 m3

  #3909 merged Aug 21, 2025

• perf: Optimize some styles

  #3907 merged Aug 21, 2025

• fix: Resource authorization

  #3906 merged Aug 21, 2025

• feat: Add connection pool connection check parameters

  #3905 merged Aug 21, 2025

• perf: Optimize packaging

  #3903 merged Aug 20, 2025

• feat: Vllm whisper model

  #3901 merged Aug 20, 2025

• fix: Internationalization Error

  #3900 merged Aug 20, 2025

• fix: Internationalization Error

  #3899 merged Aug 20, 2025

• feat: chat upload input style

  #3898 merged Aug 20, 2025

• fix: Some web pages are unable to be crawled

  #3897 merged Aug 20, 2025

• feat: AI dialogue nodes support calling tools configured in the system

  #3896 merged Aug 20, 2025

• fix: Please parse the image content. This needs to be optimized to remove the automatically occurring problems. #3837

  #3895 merged Aug 20, 2025

• fix: AI dialogue node calls MCP, AI's reply content is included in the code block of MCP call result, markdown parsing error #3846

  #3894 merged Aug 20, 2025

• fix: The website knowledge base lacks comprehensive methods for determining files and folders, resulting in ineffective access to links within the final document #2935

  #3893 merged Aug 20, 2025

• fix: Fix form style issues

  #3892 merged Aug 20, 2025

• fix: Advanced Layout Canvas Editing - Special Editing Method will connect the exit of the current node to the exit of subsequent nodes when adding them

  #3891 merged Aug 20, 2025

• perf: Optimize front-end code

  #3888 merged Aug 19, 2025

• feat: Add create MCP function

  #3885 merged Aug 19, 2025

• feat: Qwen asr speech recognition

  #3882 merged Aug 19, 2025

• fix: The issue of prologue in the conversation log

  #3880 merged Aug 19, 2025

• feat: Form nodes support file upload and multi-line text

  #3879 merged Aug 18, 2025

• feat: Front end permissions for resource authorization

  #3878 merged Aug 18, 2025

• feat: Backend permissions for resource authorization

  #3875 merged Aug 18, 2025

• fix: Form collection, multiple dragging and dropping of parameters failed(#3745)

  #3874 merged Aug 18, 2025

• perf: Segmented search function in the knowledge base, clearing search content when switching search keywords(#3867)

  #3872 merged Aug 18, 2025

• feat: Resource authorization permission

  #3873 merged Aug 18, 2025

• perf: Upgrade markdown editor version(#3322)

  #3871 merged Aug 18, 2025

• feat: Qwen2.5-omni-7b full modal speech recognition

  #3870 merged Aug 18, 2025

• fix: Upgrading psycopg

  #3869 merged Aug 18, 2025

• feat: Workflow form nodes support reference assignment

  #3866 merged Aug 15, 2025

• feat: Interface with Qwen Omni speech to text model

  #3865 merged Aug 15, 2025

• feat: Workplace resource authorization

  #3864 merged Aug 15, 2025

• fix(ui): markdown unordered list does not show correctly

  #3857 merged Aug 15, 2025

• feat: System resource authorization function

  #3861 merged Aug 15, 2025

26 Issues closed by 9 people

• [Bug] The web site type knowledge base, incomplete methods for judging files and folders, resulting in invalid link access in the final document

  #2935 closed Aug 21, 2025

• [Bug] When the answer involves mcp calls, the AI analysis results are displayed in mcp_tool calls results

  #3846 closed Aug 21, 2025

• [Bug]【API Interface】Hit the test list interface. After passing in all parameters, the response returns [Query text] This field cannot be null.

  #3808 closed Aug 21, 2025

• [Feature] v1 Please parse the image content. This needs to be optimized to remove the automatically occurring problems.

  #3837 closed Aug 21, 2025

• [Bug] [Advanced Orchestration] Select the browser to play, click to play on the Q&A page, and then when it is not finished playing, click Pause, and then click to play again, and it will not respond. You can only choose to refresh the browser to play.

  #3769 closed Aug 21, 2025

• [Bug] After the number of knowledge bases increases, the search time increases exponentially, resulting in an increase in waiting time when answering

  #3707 closed Aug 21, 2025

• [Bug] The first drag and drop of form collection parameters is successful, but subsequent drag and drop fails

  #3745 closed Aug 21, 2025

• Quick start failed - MacOS

  #2171 closed Aug 20, 2025

• [Feature] Hope the node adds traversal nodes

  #2385 closed Aug 20, 2025

• [Feature] In the knowledge base search, fields such as the number of documents related to the knowledge base, document name, etc. are referenced in the orchestration.

  #3517 closed Aug 20, 2025

• [FEATURE] 知识库文档caj格式支持

  #751 closed Aug 20, 2025

• [FEATURE] nginx反响代理细化

  #1155 closed Aug 20, 2025

• [FEATURE] ISV厂商建议MaxKB对接天猫 千牛和 京东的东东 客服。

  #1360 closed Aug 20, 2025

• [FEATURE]【知识库】问题支持分类管理

  #120 closed Aug 20, 2025

• [FEATURE]希望可以支持RWKV-Runner项目

  #420 closed Aug 20, 2025

• [Bug] The file URL is provided in the known information, and the final jump link will become the MK's own link.

  #3863 closed Aug 20, 2025

• [Feature] I hope to create documents online when uploading the knowledge base and support markdown format

  #3887 closed Aug 20, 2025

• [Feature] Hope that the file upload component can be added to the workflow

  #3884 closed Aug 20, 2025

• [Feature] ai module supports automatic retrieval rather than dependent module

  #3881 closed Aug 20, 2025

• [Bug] Application Import Error: when I try to import an application which is exported from MaxKB v1.10.8 into MaxKB v2.0.2, there is the error"不支持的文件格式"

  #3889 closed Aug 20, 2025

• [Bug] v1 uses the parameter question, and start asking questions in the demonstration, there is no problem at this time. When ai answers, the session is interrupted and created again, an error will be reported

  #3886 closed Aug 19, 2025

• [Feature] maxkb Want to enter by user, click [Start conversation] and directly contact and send AI Q&A Q&A

  #3876 closed Aug 19, 2025

• [Bug] "Dooring with Baidu Qianfan" failed

  #3862 closed Aug 18, 2025

• [Bug] Problem with editing model parameters

  #3859 closed Aug 15, 2025

• [Bug] Xinference TTS API调用报错

  #3796 closed Aug 15, 2025

• [Feature] mk's form filling must support width setting, otherwise it will be too narrow. The style is too ugly when I want to fill in long text

  #3008 closed Aug 15, 2025

6 Issues opened by 5 people

• [Bug] [(Enterprise Edition) Application] Identity authentication enables password verification, click to go to the conversation, and you can directly talk, and there is no pop-up of the password input box to enter the password.

  #3902 opened Aug 20, 2025

• [Bug] The latest version of maxkb test AI dialogue component deployed by docker is required to wait several minutes to finish the conversation after opening the MCP option.

  #3883 opened Aug 19, 2025

• [Feature] When uploading a file for multi-modal questions, the default display will be displayed after uploading the file. It is recommended to cancel the default information or edit the default information.

  #3877 opened Aug 18, 2025

• [Bug] [Knowledge Base Issue] Added multiple identical questions to segmentation, only one is displayed on the interface, but the number of segments associated with the "Problem" page shows 3, and click 3 to enter only one segment.

  #3868 opened Aug 18, 2025

• [Bug] Knowledge Base searches by title. The searched content does not exist in the title, but it exists in the knowledge base. When switching to search content, the content cannot be searched.

  #3867 opened Aug 15, 2025

• [Bug] Variable assignment is cleared

  #3860 opened Aug 15, 2025

10 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• [Bug] The hit test input box of the knowledge base after expired license exceeds the screen

  #3852 commented on Aug 15, 2025 • 0 new comments

• [Feature] Supports docking with enterprise WeChat intelligent robots

  #3751 commented on Aug 15, 2025 • 0 new comments

• [Bug] Community Edition Permission Management Issues / 权限管理问题

  #3853 commented on Aug 19, 2025 • 0 new comments

• [Feature] Function component supports multi-parameter output, especially supports output files to provide user downloads (generated output of analogy model)

  #3824 commented on Aug 20, 2025 • 0 new comments

• [Feature] Form node type supports file upload

  #2482 commented on Aug 20, 2025 • 0 new comments

• [Feature] Alibaba Cloud Bailian's speech recognition model needs to support Asr and Paraformer-v2

  #3560 commented on Aug 20, 2025 • 0 new comments

• [Bug] Alibaba's full-modal model of qwen-omni-turbo and qwen2.5-omni-7b, reported an error when added as speech recognition or synthesis model

  #3849 commented on Aug 20, 2025 • 0 new comments

• [Feature] You can customize the default login method

  #2895 commented on Aug 20, 2025 • 0 new comments

• [Feature] Dialogue log details Default top dialogue display

  #2957 commented on Aug 20, 2025 • 0 new comments

• [Feature] Add reranker model support to vllm

  #2938 commented on Aug 20, 2025 • 0 new comments