> For the complete documentation index, see [llms.txt](https://docs.maiagent.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.maiagent.ai/release-note/en/readme.md).
# Changelog
MaiAgent platform updates and new feature releases
Past releases are archived by year and month under "Historical Versions" on the left.
## v2026.10.19
***
### 📌 Highlights
* **\[Shared] GPT-5.6 Sol via direct OpenAI stays available**: The GPT-5.6 Sol model connected directly to OpenAI is no longer retired by a scheduled job after deployment, so AI assistants using it are not automatically moved back to the Azure endpoint.
* **\[Shared] Token-by-token streaming for OpenAI models on Bedrock**: AI assistants using GPT-6.1 Sol and GPT-5.6 Sol on Bedrock now show answers as they are generated, instead of waiting for the full response before displaying it all at once.
* **\[Work] Dedicated panel for MaiGPT document translation**: MaiGPT now presents document translations in a dedicated translation card, translating paragraph by paragraph and showing the source file details along with the source and target languages. [User Guide](https://docs.maiagent.ai/work/maigpt)
* **\[Build] Daily and monthly Credits spending limits for API keys**: Administrators can set a daily and a monthly Credits spending limit on each API key, and calls are rejected once a limit is exceeded. [User Guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Serve] Internal notes in conversations**: Colleagues can leave notes in a conversation that only internal members can see, synced in real time to other colleagues who are online. [User Guide](https://docs.maiagent.ai/serve/daily-operations/reply)
* **\[Shared] Vision Parser Economy tier**: Knowledge base files can use the Vision Parser Economy tier, which is billed per page, and the parser list shows the price per page. [User Guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Build] Assistant tool loading mode: Auto, All, or Dynamic Search**: In settings, an AI assistant can choose to load tools automatically, load all of them, or find them through dynamic search, and tools found by search are kept for the whole conversation. [User Guide](https://docs.maiagent.ai/build/agent-builder)
* **\[Work] Schedules auto-disable after consecutive failures, and short intervals require a stop condition**: A schedule is disabled automatically, and its creator notified, when consecutive failures reach the limit or the end date arrives. Schedules that run more often than once an hour must have a stop condition. [User Guide](https://docs.maiagent.ai/work/overview)
* **\[Serve] Automatic conversation splitting and carry-over summaries on connected messaging platforms**: Conversations on platforms such as LINE and Messenger are split automatically once resolved or idle, and the new conversation starts with a carry-over summary. [User Guide](https://docs.maiagent.ai/serve/daily-operations/reply)
* **\[Work] AI assistant file changes run only after user confirmation**: In organizations that have not enabled skipping confirmation, MaiGPT first lists its planned file changes and only creates folders, renames, and moves files after the user confirms.
### 🗂️ Full Update (by Product Line)
Work (44 items)
#### ✨ New Features
* **\[Work] Dedicated panel for MaiGPT document translation**: MaiGPT now presents document translations in a dedicated translation card, translating paragraph by paragraph and showing the source file details along with the source and target languages. [User Guide](https://docs.maiagent.ai/work/maigpt)
* **\[Work] Schedules auto-disable after consecutive failures, and short intervals require a stop condition**: A schedule is disabled automatically, and its creator notified, when consecutive failures reach the limit or the end date arrives. Schedules that run more often than once an hour must have a stop condition. [User Guide](https://docs.maiagent.ai/work/overview)
* **\[Work] MaiGPT organization sharing**: A conversation shared with the organization follows the original conversation, and its creator can change who it is shared with and keep chatting at any time. [User Guide](https://docs.maiagent.ai/work/maigpt)
* **\[Work] Desktop Cowork local browser tool integration**: MaiGPT's Desktop Cowork can use local browser tools, and the cloud browser is no longer offered alongside it when a local browser is available. [User Guide](https://docs.maiagent.ai/work/maigpt)
* **\[Work] Export translations as Word files that keep the original formatting**: After a Word document is translated, you can download the translated version as a Word file that keeps the original formatting. [User Guide](https://docs.maiagent.ai/work/maigpt)
* **\[Work] Layout-preserving PDF export can reuse Canvas translations**: When exporting a PDF that keeps the original layout, you can choose to reuse the content already translated and edited in Canvas. [User Guide](https://docs.maiagent.ai/work/maigpt)
* **\[Work] Personal secretary management and admin console**: Organizations with the personal secretary feature enabled can manage personal secretaries in the admin console, and the personal conversation list also shows the user's own Work conversations. [User Guide](https://docs.maiagent.ai/work/maigpt)
* **\[Work] Quality validation before document translation export**: Before a translated document is exported, its text and layout quality are checked, and files that fail are not offered for download as successful results. [User Guide](https://docs.maiagent.ai/work/maigpt)
* **\[Work] In-place translation of text in images embedded in documents**: Text inside images embedded in PDF and Word documents can now be translated in place. [User Guide](https://docs.maiagent.ai/work/maigpt)
* **\[Work] Resumable uploads in the file library**: The file library adds a resumable upload mechanism, so an interrupted upload can continue from where it stopped, with integrity verification. [User Guide](https://docs.maiagent.ai/work/file-library)
* **\[Work] Meeting record live speech engine details and live translation model choice**: The live speech engine for meeting records shows its country of origin, data location, and supported spoken languages, and live translation lets you pick a model from three tiers. [User Guide](https://docs.maiagent.ai/work/meeting-records)
#### 🚀 Core Performance Improvements
* **\[Work] Prompt cache continues after a Canvas handoff**: After a handoff, the Canvas assistant reuses the main assistant's tools and system instructions so the prompt cache carries over, cutting credit consumption by about 4.3%.
* **\[Work] Shorter wait for responses when AI Memory is labeled**: The AI Memory location estimate now runs outside the main model call, and the limit on how much memory is included uses a new counting method.
* **\[Work] Faster conversation visibility queries**: When members view the conversation list, visibility is now determined with a more efficient query, so large organizations load faster.
* **\[Work] Unified tool loading mode for MaiGPT**: The tool loading mode of every MaiGPT is now decided by the platform template, and legacy MaiGPTs no longer load all tools on every turn.
#### 😊 User Experience and Interface Improvements
* **\[Work] File library browse page shows version history and usage**: The file library browse page adds library-wide version history, used storage, text previews, and file creator information.
* **\[Work] Handling and notification for schedules with insufficient credits**: A schedule that hits insufficient credits no longer counts toward its run total, and its creator is notified when it is disabled for insufficient credits or for reaching its run limit.
* **\[Work] MaiGPT workspace removes the 20-file limit per conversation**: The MaiGPT workspace can now read files in a single conversation without the 20-file cap.
* **\[Work] Slides skill accepts typed answers for attachment and template questions**: When the AI slides skill asks what an attachment is for and which template to use, users can answer by typing text directly.
* **\[Work] More accurate live translation of speech-recognition fragments**: Meeting record live translation now treats its input as speech-recognition fragments, cutting the average number of mistranslated segments from about 19 to about 15.
#### 🔐 Security and Governance Enhancements
* **\[Work] Default roles no longer carry file library permissions**: File library access is now granted through explicitly assigned custom roles, and file library permissions on default roles no longer take effect.
#### 🛠️ Bug Fixes
* **\[Work] AI assistant file changes run only after user confirmation**: In organizations that have not enabled skipping confirmation, MaiGPT first lists its planned file changes and only creates folders, renames, and moves files after the user confirms.
* **\[Work] Repeated content in audio transcripts is collapsed automatically**: After an uploaded audio file is recognized, the transcript no longer contains hundreds of repetitions of the same words, and a repeated phrase is kept at most 3 times.
* **\[Work] Audio transcription moves to Groq because of the OpenAI model retirement**: WhisperOpenAI audio transcription for knowledge bases and meeting records is now handled by a Groq model, and segment timestamps are kept.
* **\[Work] Stability fix for user memory extraction**: When MaiGPT memory organizes facts about a user, it no longer mistakenly answers instructions found in messages, and extraction results are saved normally.
* **\[Work] Replies save correctly when searching conversation messages with a date range**: After asking the assistant to search conversation messages by date range, the answer is saved to the conversation history normally and does not disappear on refresh.
* **\[Work] MaiGPT map canvas no longer appears as raw text**: When you ask MaiGPT to mark locations on a map, the map is shown in the side panel instead of a block of raw data appearing in the conversation.
* **\[Work] Agents in the file workspace can create folders and complete rename and move together**: MaiGPT and file workspace Agents can create new folders, and renaming and moving the same file now happen together.
* **\[Work] Slide export images support the SSL relaxed domain list**: Image retrieval during slide export now applies the SSL relaxed domain list, so on-premises sites with TWCA certificates can export correctly.
* **\[Work] Missing download-link files now suggest available files**: When the AI assistant requests a download link for a file that does not exist, it sees the list of downloadable files in that conversation and can choose another one.
* **\[Work] AI Slides are built to the page count the user types**: When AI Slides asks for the number of pages, a number the user types in is used directly as the page count.
* **\[Work] PDF translation supports PDFs printed from a browser**: PDFs printed from a browser can now produce a layout-preserving bilingual PDF, and a clear reason is shown when a file cannot be processed.
* **\[Work] PDF translation reads multi-column layouts in order**: Multi-column PDFs are translated column by column, and sentences split across line breaks are merged into complete paragraphs.
* **\[Work] Canvas agent default prompt adds Mermaid line-break rules**: The line-break rules for node text in Mermaid flowchart canvases are now part of the default prompt; clear the custom value in the admin console to apply the latest rules.
* **\[Work] AI Slides verifies images as each page is written**: When AI Slides writes a page containing generated images, it checks on the spot whether the images are displayed, and reports clearly if they are not instead of reporting completion.
* **\[Work] Image-to-video can use images uploaded in the conversation**: Image-to-video can use images uploaded in the conversation directly, and they are no longer blocked as external URLs.
* **\[Work] Tools with a runtime parameter no longer cause reply failures**: Tools with an execution-environment parameter, such as Desktop Run Script, now run and finish normally instead of showing "The system is temporarily unable to process this request."
* **\[Work] MaiGPT desktop browser tools stay enabled after deployment**: The Desktop local browser tools are enabled and bound to MaiGPT automatically after deployment, with no need to turn them on manually.
* **\[Work] Fixed Desktop tools being rejected because of optional parameters**: Desktop tools with optional parameters are no longer rejected, and parameters the model left blank are not forwarded.
* **\[Work] Fixed being unable to rename a meeting right after a long recording ends**: Meeting names can be changed immediately after a long recording ends, even while it is still being converted, and edits no longer overwrite the conversion result.
* **\[Work] PDF translation restores the Layout tab**: The Layout tab in the PDF translation canvas is shown again, so paragraph positions can be viewed.
* **\[Work] File library checks content before applying pending suggestions**: Applying a pending file library suggestion now confirms the content has not changed, so content the user did not preview is never applied.
* **\[Work] Fixed file size of exported translated PDFs**: When document translation exports a PDF in the original layout, the number of image labels no longer multiplies the file size.
* **\[Work] Images generated before AI Slides are created can be used in Insert Image**: Images that MaiGPT generated before creating a slide deck now also appear in the Insert Image drawer and the deck's folder.
Serve (37 items)
#### ✨ New Features
* **\[Serve] Internal notes in conversations**: Colleagues can leave notes in a conversation that only internal members can see, synced in real time to other colleagues who are online. [User Guide](https://docs.maiagent.ai/serve/daily-operations/reply)
* **\[Serve] Automatic conversation splitting and carry-over summaries on connected messaging platforms**: Conversations on platforms such as LINE and Messenger are split automatically once resolved or idle, and the new conversation starts with a carry-over summary. [User Guide](https://docs.maiagent.ai/serve/daily-operations/reply)
* **\[Serve] Assistant sign-in sources support organization sign-in settings and AD name validation**: Messaging platform sign-in settings gain a new "MaiAgent organization sign-in settings" source, and the SSO name for AD is validated before it is saved. [User Guide](https://docs.maiagent.ai/serve/daily-operations/reply)
* **\[Serve] Automatic closing of idle live-agent conversations**: Messaging platforms can enable idle auto-close, so a conversation closes automatically when the customer does not reply within the configured number of minutes after the agent responds. [User Guide](https://docs.maiagent.ai/build/conversation-settings/human-handoff)
* **\[Serve] Messenger and Instagram icebreaker questions and persistent menu**: Messenger and Instagram messaging platforms can set icebreaker questions per language and sync the assistant's quick menu as a persistent menu. [User Guide](https://docs.maiagent.ai/build/choose-channel)
* **\[Serve] Voice support adds gpt-realtime-whisper**: Voice assistants gain OpenAI gpt-realtime-whisper streaming speech recognition as a new option. [User Guide](https://docs.maiagent.ai/serve/voice)
* **\[Serve] Messaging platforms can replace the Access Token**: People with permission to create messaging platforms can replace the credentials of LINE, Telegram, Viber, Slack, and Microsoft Teams messaging platforms directly, without deleting and recreating them. [User Guide](https://docs.maiagent.ai/serve/daily-operations/reply)
* **\[Serve] Web chat location request timing setting**: Messaging platforms gain a location request timing setting, so a visitor's location can be requested only when the AI asks with a question card. [User Guide](https://docs.maiagent.ai/serve/daily-operations/reply)
* **\[Serve] The conversation API accepts an external user ID directly**: When sending a message through the conversation API, an external user ID can be included, and the system automatically finds or creates the matching contact. [User Guide](https://docs.maiagent.ai/org/developer)
* **\[Serve] LINE replies can include a sticker**: With this enabled on a LINE messaging platform, an AI reply can include one LINE sticker along with the text. [User Guide](https://docs.maiagent.ai/build/choose-channel)
* **\[Serve] Messenger and Instagram assistant cards are sent as native cards**: Card messages from the assistant are now presented as Meta native cards on Messenger and Instagram. [User Guide](https://docs.maiagent.ai/build/choose-channel)
#### 🚀 Core Performance Improvements
* **\[Serve] Faster unread-count loading in the inbox**: When non-owner members open a large inbox such as MaiGPT, unread counts now use an indexed query, so they no longer occasionally wait over ten seconds.
* **\[Serve] Faster loading of the All Conversations list**: The All Conversations list now uses an index that only covers conversations with incoming messages, so it no longer times out while passing through large numbers of empty conversations.
* **\[Serve] Faster voice call setup**: Voice support now picks up a pre-created room when it requests a call connection, which shortens the wait.
#### 😊 User Experience and Interface Improvements
* **\[Serve] Conversation analysis summary written automatically on human handoff**: When a conversation is handed off to a human and conversation analysis is enabled on the platform, the AI summary now appears in the conversation details.
* **\[Serve] Fixed error codes in platform API error responses**: The send-message API now returns a fixed error code when a contact error occurs, so callers can tell what went wrong.
#### 🔐 Security and Governance Enhancements
* **\[Serve] Webhook and Slack credentials stored encrypted**: The Secret Token for message webhooks and the Bot Token and Signing Secret for Slack are now stored encrypted.
* **\[Serve] Handling of disabled Web Chat IDs at the login entry point**: At the login entry point, a Web Chat ID disabled by another organization now gives the same result as an ID that does not exist.
* **\[Serve] Telegram channel verifies the push source**: The Telegram channel now verifies where push requests come from, and forged requests are rejected.
#### 🛠️ Bug Fixes
* **\[Serve] File uploads on messaging platforms download in the background and report failures**: Files that end users send on messaging platforms are now downloaded in the background with a size limit. If a download fails, the user is notified and other messages are no longer lost along with it.
* **\[Serve] No more partial data when a Web Chat message fails to send**: When Web Chat sends a message, the related data is now committed together, and if any part fails, everything is rolled back instead of leaving an incomplete state.
* **\[Serve] Image URLs without a file extension are now sent as images**: GitBook screenshots referenced in replies now appear as images on LINE, Messenger, Instagram, WhatsApp, and Team+.
* **\[Serve] Telegram typing indicator and LINE WORKS bot rename handling**: Telegram groups now consistently show the typing indicator when certain documents are received, and LINE WORKS bots are recognized automatically under their new name after a rename.
* **\[Serve] LINE sticker tags no longer leak**: Sticker tags in AI assistant replies no longer appear as raw text on any conversation channel.
* **\[Serve] Voice and video from channels play directly**: Voice messages and videos received from channels can now be played directly in the conversation.
* **\[Serve] Fixed LINE emoji placeholder text**: Emoji placeholder text in LINE messages is no longer mistaken for a person's name by the AI.
* **\[Serve] Deep Research keeps original file names when analyzing photos**: During Deep Research, the internal assistant now sees the original file name of every photo a customer uploaded, such as "receipt.png".
* **\[Serve] Gemini abnormal termination is retried automatically, so LINE no longer occasionally goes unanswered**: When Gemini ends with an empty response and an unrecognized finish code, the request is now resent automatically, and LINE conversations no longer occasionally get no response.
* **\[Serve] Abnormal attachment IDs are no longer silently dropped when sending messages**: When a message is sent with a missing or malformed attachment ID, the attachment is no longer silently discarded and lost.
* **\[Serve] Full-width ＠ mentions in groups also trigger the assistant**: Mentioning the assistant with a full-width ＠ followed by its name in LINE, LINE WORKS, and Telegram groups is now treated as a valid mention.
* **\[Serve] Telegram and other channel replies no longer include code and raw data**: AI replies on Telegram and other channels no longer include the code and raw data that the assistant wrote into files.
* **\[Serve] AI replies are no longer interrupted by background machine scale-down**: AI replies on Telegram, LINE, and other channels are no longer cut off when background machines scale down, so users receive their replies normally.
* **\[Serve] Reset commands on all channels only start a new conversation**: Entering a reset command on any conversation channel now only opens a new conversation, without passing it to the AI for a reply or consuming credits.
* **\[Serve] Fixed blank conversation record panel for LINE, Slack, and Teams**: Conversations from the LINE, Slack, and Teams channels now display their content in the admin "Full Conversation Record" panel.
* **\[Serve] Fixed file-write content reaching Telegram under rewriting hooks and masking**: When an organization has a rewriting Hook or content masking enabled, Telegram and other channels no longer receive content written to files, and memory now stores a file pointer instead.
* **\[Serve] Email channel retries daily after a login rejection**: After a login is rejected, the Email channel now retries automatically every day instead of staying down permanently.
* **\[Serve] Claude 5.5 can be required to attach text with Flex cards**: When "Require text with cards" is enabled, Claude 5.5 Sonnet and Opus now retry to obtain reply text instead of sending only the card.
Build (76 items)
#### ✨ New Features
* **\[Build] Daily and monthly Credits spending limits for API keys**: Administrators can set a daily and a monthly Credits spending limit on each API key, and calls are rejected once a limit is exceeded. [User Guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Build] Assistant tool loading mode: Auto, Load All, or Dynamic Search**: In AI assistant settings, you can choose Auto, Load All, or Dynamic Search for tool loading, and tools found through search stay loaded for the rest of the conversation. [User Guide](https://docs.maiagent.ai/build/agent-builder)
* **\[Build] AI assistants can specify their own question rewriting model**: The advanced settings of an AI assistant now let you pick a question rewriting model, and the system setting is used when none is selected. [User Guide](https://docs.maiagent.ai/build/agent-builder)
* **\[Build] Files returned by MCP tools are saved as conversation attachments**: Images, audio, and documents returned by MCP tools are automatically saved as conversation attachments that the AI can read directly. [User Guide](https://docs.maiagent.ai/build/tools)
* **\[Build] Agent health check**: You can run a health check on an AI assistant to list problems with its instructions, skills, and model settings, along with suggested fixes. [User Guide](https://docs.maiagent.ai/build/agent-builder)
* **\[Build] Files downloaded by tools are saved as conversation attachments automatically**: Files returned by API tools or MCP tools (such as PDFs, Excel files, and images) are saved as conversation attachments, so the assistant can read their contents directly. [User Guide](https://docs.maiagent.ai/build/tools)
* **\[Build] Duplicate file detection for knowledge base uploads**: Uploading a file with identical content to the same knowledge base is rejected by the system. [User Guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Build] OCR on scanned pages enabled by default, with per-page billing for MaiAgent Parser scanned pages**: Document parsing now enables OCR references for scanned pages by default, and MaiAgent Parser adds OCR to scanned pages and bills them per page. [User Guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Build] MaiAgent Parser supports HTML files**: MaiAgent Parser can now parse .html and .htm files, for both knowledge base uploads and conversation attachments. [User Guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Build] Auto-detection switched to per-page structural classification**: Auto-detection for knowledge base uploads now classifies each page individually, scanned files go straight to Vision, and shared images and logos are no longer misjudged. [User Guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Build] AI assistant defaults**: Organizations can set defaults for AI assistants that are applied automatically to new assistants and can be enforced on existing ones. [User Guide](https://docs.maiagent.ai/build/agent-builder)
* **\[Build] vLLM and dots.mocr options added to the on-premises trial Parser**: Knowledge base parsing can use an OpenAI-compatible (vLLM) endpoint and dots.mocr to recognize scanned pages, and the system status page shows all three endpoints, where a dormant endpoint is not counted as a failure. [User Guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Build] On-premises trial parsers**: When uploading files to a knowledge base, you can choose OCR (GPU) and two Vision parsers, Gemma 4 and Qwen 3.8, to try the parsing quality of the on-premises offering first. [User Guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Build] Agent health check batch ranking and one-click fixes**: Administrators can run a health check on all editable AI assistants in one batch, with results ranked by severity and suggested fixes that can be applied. [User Guide](https://docs.maiagent.ai/build/agent-builder)
#### 🚀 Core Performance Improvements
* **\[Build] Faster code sandbox execution**: The code execution tool reuses the sandbox within a single reply turn instead of starting and stopping it on every call.
* **\[Build] Skill resource bundle caching**: Skills with resource files share a cache by version, so each conversation no longer downloads and repackages them.
* **\[Build] MaiAgent Parser (Offline) runs OCR only on blocks that need it**: When MaiAgent Parser (Offline) parses digital PDFs, tables and shaded blocks that already contain text are read directly instead of being recognized again.
* **\[Build] Browser panel frames are sent only to people who are watching**: Browser Tool panel frames are sent only to users who are viewing the panel, limited to at most 5 frames per second.
* **\[Build] MCP tools share one connection within a conversation turn**: When the same MCP server is called multiple times in one conversation turn, a single connection is shared, reducing repeated handshake waits.
* **\[Build] Better prompt cache hits for Claude assistants with many tool calls**: For Claude-based assistants in conversations with many tool turns, the next turn can reuse the cache up to the previous question instead of rewriting the entire history.
* **\[Build] OpenAI/Azure standard replies exclude knowledge base retrieval content from the cache prefix**: For standard replies on OpenAI and Azure, knowledge base retrieval content is moved out of the cache prefix to improve prompt cache hits.
* **\[Build] Test dataset evaluation saves re-query costs**: Test dataset evaluation uses JSON replies that arrive as plain text directly, reducing Sonnet 5.5 evaluation cost by about 37%.
* **\[Build] Hook sandbox pre-warming**: Platform administrators can turn on Hook sandbox pre-warming in system settings to shorten the wait for each Hook check (off by default).
* **\[Build] Faster Credit totals in the AgentOps text assistant overview**: The Credit total in the AgentOps text assistant overview now queries only the organization's own transactions, so large organizations can see the value too.
* **\[Build] AI Gateway connection tests report upstream overload quickly**: When a custom-model connection test in AI Gateway hits an upstream overload, the cause is reported in about 1 second instead of waiting two minutes.
* **\[Build] Short PDFs no longer queue during Vision parsing**: When multiple short PDFs are uploaded at once with Vision Parser, they can be parsed simultaneously instead of queuing because of an overestimated memory requirement.
* **\[Build] Faster content safety Hook and human-handoff decisions**: The content safety Hook now sends multiple segments in one request, and both Hook decisions and human-handoff decisions explicitly turn off thinking to respond faster.
#### 😊 User Experience and Interface Improvements
* **\[Build] AgentOps can list only organization-created assistants**: The AgentOps assistant list can show only assistants created by the organization and hide the official built-in assistants.
* **\[Build] Tool search keeps only long-tail tools**: Assistants with tool search enabled now get platform features and commonly used tools directly, and only long-tail tools need to be searched.
* **\[Build] The knowledge base retrieval tool supports multiple queries at once**: The knowledge base retrieval tool can handle multiple queries in one call, with results grouped by query.
* **\[Build] Vision parsing attaches OCR references for scanned pages**: The SaaS Vision parser first obtains Azure OCR results for scanned pages for the model to cross-check, improving recognition accuracy for proper nouns.
* **\[Build] The reason OCR was not run on scanned pages is now visible**: Knowledge base files show why OCR was not run on scanned pages, such as insufficient credits or an OCR service that is not configured.
* **\[Build] Skill reference file results are no longer truncated**: When an assistant reads skill reference files, the results are no longer truncated and the full content is provided to the model.
* **\[Build] Knowledge base files show the decision record of the auto-detected parser**: Knowledge base file information now shows the parser selected by auto-detection and the reason for the decision, and records any later change of parser made by the user.
* **\[Build] Hook execution logs show external call durations**: Hook execution logs and exported files now list the name, duration, and success or failure of each external call.
* **\[Build] Parser menu cleanup**: The parser menu shown when uploading files has a new order, MaiAgent Vision Parser is renamed, the CPU version is listed last, and retired options are removed.
* **\[Build] Parser menu shows pricing basis and reference prices**: The parser menu adds an explanation of the pricing basis, and Vision and on-premises trial parsers show a reference price when no measured data is available yet.
#### 🔐 Security and Governance Enhancements
* **\[Build] Calls rejected when a self-hosted model credential is disabled**: When the credential of a self-hosted model is disabled or missing, the call is rejected instead of falling back to the platform key.
* **\[Build] Attachment ownership check also covers URL parameters of MCP tools**: If a URL-type parameter of an MCP tool points to an attachment from another conversation or organization, that call is blocked.
* **\[Build] Protection against path rewriting that bypasses MCP attachment URL checks**: The attachment ownership check for MCP tools now blocks URLs that use path rewriting to point to attachments in other users' conversations.
* **\[Build] Bring-your-own vector database keys and passwords stored encrypted**: API keys for bring-your-own vector databases and the connection passwords for OpenSearch, Oracle, and SynxDB are now stored in encrypted form.
* **\[Build] High-cost models must be explicitly selected on the key**: For designated organizations, an API key must explicitly select high-cost models such as Opus in its available models before it can call them.
* **\[Build] Assistant database connection settings stored encrypted**: The database connection settings of an assistant are now stored in encrypted form.
#### 🛠️ Bug Fixes
* **\[Build] Agent replies that were cut off can be resumed**: After an Agent reply is cut off by a timeout, the user can say "continue" to resume from the previous progress.
* **\[Build] Fix for blank wrap-up after delegation**: After an AI assistant team delegates a task, the answer from the delegated target is relayed even if the source assistant's wrap-up is blank.
* **\[Build] Fix for AgentOps viewing colleagues' MaiGPT conversations**: AgentOps monitors who open a colleague's MaiGPT conversation from the record details no longer get a page-not-found error.
* **\[Build] Progress summary requested when the tool call limit is reached**: When the tool call limit is used up and the model has left no text, the model is asked once more to explain the current progress.
* **\[Build] Vision parsing attaches only images actually shown on the page**: When the Vision parser processes a PDF exported from a presentation, each page now references only the images drawn on that page and no longer mixes in images from other pages.
* **\[Build] Gemini no longer fails to reply when a tool name starts with a digit**: When an API tool name starts with a digit, assistants that use Gemini no longer fail the entire reply.
* **\[Build] Prompt injection protection no longer blocks skill expansion and tool search**: After an organization enables prompt injection protection for Agent assistants, assistants can still expand the organization's own skills and search for tools normally.
* **\[Build] Web fetch tool validates multiple quotes before returning the whole page**: The web fetch tool first validates multiple quotes, and only returns the whole page content to the model when validation fails.
* **\[Build] Wrap-up message is no longer blank when the tool use limit is reached**: When an assistant uses up its tool-use count for the round and the model leaves no visible content, a default wrap-up message is shown instead.
* **\[Build] No more 401 when refreshing the tool list while editing an MCP tool**: When editing an MCP tool, clicking "Refresh" without changing the headers no longer produces an HTTP 401 error.
* **\[Build] Conversation attachments using the Offline parser are now routed to a worker**: When an attachment uploaded in a conversation uses MaiAgent Parser (Offline), it is now actually parsed by that parser instead of being silently replaced by another one.
* **\[Build] File retrieval and concurrent file reading no longer interfere with each other**: When the AI performs retrieval restricted to specific files while reading other files, the reads are no longer wrongly narrowed in scope and fail.
* **\[Build] Failed tool search now suggests using a skill or an existing tool**: When a tool search comes up empty, the AI is prompted to expand a skill or call a tool it already has, instead of wrongly replying that the feature does not exist.
* **\[Build] Conversations created by an API key can use MCP tools under the caller's identity**: Conversations invoked with an API key can now use MCP tools as the member who owns that key, instead of always failing authorization.
* **\[Build] Pages blocked by Vision content filtering are filled in from the text layer**: When the Vision Parser encounters a page blocked by content filtering, it now fills in the page using the text layer content instead of dropping the whole page.
* **\[Build] Team delegation relay pauses only when the user sees the file panel**: The relay of a team delegation pauses only when the user actually sees the file panel; in all other cases the full reply is kept.
* **\[Build] Automatic resend on transient Azure tool-argument streaming errors**: Assistants using Azure models now automatically resend when a transient streaming error occurs during a long tool workflow, instead of being interrupted.
* **\[Build] Skill instructions stay whole under size-based truncation**: For assistants with size-based truncation enabled, skill instructions are kept in full after a skill is expanded and are no longer truncated.
* **\[Build] Evaluation custom metrics fall back to lazy generation on empty steps**: When the judge model returns empty steps for a custom metric of an evaluation dataset, the steps are now generated lazily instead of returning the whole batch.
* **\[Build] Message search and segmented reading of long messages fixed**: Message search no longer fails because of regular expression errors, and messages longer than 1,000 characters can be read in full in segments.
* **\[Build] Document parsing OCR decision and billing fixes**: Document parsing now runs OCR only on genuinely scanned pages and bills correctly, and presentation files no longer trigger an extra model decision call.
* **\[Build] Database connection test times out within 10 seconds**: When adding a database connection, testing an unreachable host now shows "Connection failed" within 10 seconds instead of hanging for about two minutes.
* **\[Build] AI assistant monitoring text assistant overview no longer times out for large organizations**: AgentOps' text assistant overview now loads for large organizations too, and shows "-" instead of an error when Credit consumption cannot be calculated in time.
* **\[Build] Fixed role permissions being lost when duplicating an AI assistant**: Duplicating an AI assistant now keeps each role's original read-only, edit, and delete permissions.
* **\[Build] OCR of Traditional Chinese scans no longer mixes in Simplified characters**: After parsing with Docling and MaiAgent Parser, OCR results for Traditional Chinese scanned documents are now consistently in Traditional characters.
* **\[Build] Unresponsive external MCP servers wait at most 30 seconds**: When an external MCP server does not respond, the assistant now replies after at most 30 seconds instead of waiting 5 minutes.
* **\[Build] Replies intercepted by a Hook are no longer misreported as errors**: When a Hook intercepts an AI assistant's reply card, conversations and evaluation results are now marked as intercepted rather than reported as system errors.
* **\[Build] Tool result attribution fixed after Agent Teams handoff**: After an Agent Teams handoff, the receiving assistant's tool call results are no longer mistakenly replaced by the previous assistant's output.
* **\[Build] Testing an on-premises parser on the system status page no longer wakes a sleeping GPU**: When administrators test the trial parser on the system status page, a sleeping endpoint now shows "Sleeping" and is no longer woken up or triggers alerts.
* **\[Build] Longer cold-start wait for the trial parser**: The trial parser's cold-start wait limit is extended to 10 minutes, and the description now reflects the measured three to six minutes.
* **\[Build] Earlier masks are preserved when a Hook blocks or skips input**: When a Hook blocks or skips input, the mask applied by a previous Hook is preserved, so sensitive content is not leaked.
* **\[Build] OCR parser straightens skewed scanned pages before reading tables**: The OCR parser now straightens skewed scanned pages before reading tables, so long tables no longer end up with misaligned rows.
* **\[Build] Manual-selection-only parsers are skipped when Vision parsing falls back**: When Vision parsing fails and falls back to MaiAgent Parser, parsers reserved for manual user selection are no longer chosen.
Shared (62 items)
#### ✨ New Features
* **\[Shared] Vision Parser economy tier**: Knowledge base files can use the Vision Parser economy tier, billed per page, and the parser list shows the price per page. [User Guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Shared] Role permission governance opened to all organizations**: Role permission governance is now available to every organization, with batch permission settings on the roles page and effective permissions visible on the members page. [User Guide](https://docs.maiagent.ai/org/roles)
* **\[Shared] Added GLM-5.3 model and preserved Azure reasoning content**: The model menu now includes GLM-5.3, and reasoning models on Azure can show and carry forward their reasoning content in conversations. [User Guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Shared] Added Cohere Embedding and Reranker on Azure and Bedrock**: Embedding and Reranker models gain Cohere options from the Azure and Bedrock providers, with the same point deduction as the models of the same name. [User Guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Shared] Enabled Qwen3.8 Max**: Qwen3.8 Max is now selectable for assistants and model invocation permissions, succeeding the retired Qwen3.6 Max. [User Guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Shared] Embedding and Rerank included in organization model invocation permissions**: Organization owners can switch Embedding and Rerank models on or off in model invocation permissions, and disabled models no longer appear in the knowledge base settings menu. [User Guide](https://docs.maiagent.ai/org/overview)
* **\[Shared] Enabled Qwen3.6 27B**: The assistant model menu now includes Qwen3.6 27B, the same model shipped for on-premises deployments. [User Guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Shared] Organization owners can edit members' names within the organization**: Organization owners can change a member's name within the organization through member editing, the API, or Excel import, and every rename is recorded in the activity log. [User Guide](https://docs.maiagent.ai/org/overview)
* **\[Shared] Admin system status page (external service checks)**: The admin console adds a system status page that checks external services daily and flags the features degraded as a result. [User Guide](https://docs.maiagent.ai/org/overview)
* **\[Shared] Conversation attachments automatically choose a parser by content**: Conversation attachments now automatically pick a parser based on their content, and the conversation platform's "Enable OCR" option is removed. [User Guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Shared] Organization-level AI assistant defaults**: Organizations gain settings storage and a dedicated permission for AI assistant defaults, so administrators can read and write organization-level default settings. [User Guide](https://docs.maiagent.ai/org/overview)
* **\[Shared] Organization announcements can remind on every login**: Organization announcements support a "every login during the announcement period" reminder frequency; members can choose to stop being reminded, and owners can remind everyone again. [User Guide](https://docs.maiagent.ai/org/overview)
* **\[Shared] Model capability scores added and input processing speed removed**: The model invocation permissions page now shows a capability score for each model, and the inaccurate input processing speed is removed. [User Guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Shared] Successor models are automatically added to API key allowlists with notification when a model is retired**: When a model is retired, the system automatically adds its successor to the API key's AI Gateway allowlist, and notifies the key holder, organization owners, and support if it cannot be added. [User Guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Shared] Gemma 4 supports thinking depth**: Gemma 4 thinking can now be turned on or off in assistant settings, supported on Bedrock Mantle, OpenRouter, and self-hosted vLLM. [User Guide](https://docs.maiagent.ai/build/ai-gateway)
#### 🚀 Core Performance Improvements
* **\[Shared] Offline parser upgraded with faster scanned-document recognition**: After the MaiAgent offline parser upgrade, scanned-page recognition recall rises from 61.7% to 94.0%, and per-page processing time drops from 32.7 seconds to 7.8 seconds.
* **\[Shared] Azure and OpenAI prompt caching now hits after memory recall adjustment**: For multi-turn conversations on Azure OpenAI and OpenAI with long-term memory enabled, historical content now reliably hits the prompt cache.
* **\[Shared] Faster consumption record list loading**: The credit consumption record list for large organizations is now aggregated only once and uses an index, so the page no longer takes a long time to open.
* **\[Shared] Tool declarations loaded on demand for AI assistants**: Attachment and conversation history tools are no longer declared when the conversation has no files or does not need to look up history, saving about 2.5K tokens for 73% of Agent replies.
* **\[Shared] Faster first sandbox call in each turn**: The wait time of the sandbox isolation probe is reduced from 3 seconds to 1.5 seconds, so the first sandbox call in each turn waits 1.5 seconds less.
* **\[Shared] Faster contact usage statistics loading**: The contact pagination of usage statistics now groups by ID and queries names only for the current page, so organizations with many contacts no longer wait 3 to 22 seconds.
* **\[Shared] Faster organization used-storage query**: The query for used upload storage on the organization information page now uses a dedicated index, so it loads faster.
* **\[Shared] Pre-reply credit estimation no longer blocks replies**: When the credit estimation cache is invalidated, it is now recalculated in the background, so users no longer wait for their reply.
#### 😊 User Experience and Interface Improvements
* **\[Shared] Connector request notification emails now in Traditional Chinese**: Connector request notification emails are now sent in Traditional Chinese.
* **\[Shared] Shared space permission descriptions updated**: The descriptions of the shared space create and manage permissions now read "Create and delete shared spaces, and edit space names and authorized roles".
* **\[Shared] Invitation API expiration supports two formats**: The invitation expiration and credit quota expiration fields of the invitation API now accept both ISO 8601 and millisecond timestamps.
* **\[Shared] Model access page lists blocked models when Chinese models are disabled**: After "Disable Chinese models" is turned on, the model access page still lists Chinese platform models and marks them as blocked.
#### 🔐 Security and Governance Enhancements
* **\[Shared] OAuth password grant now goes through LDAP authentication**: In organizations with LDAP enabled, exchanging a token through the OAuth password grant now also passes LDAP authentication.
* **\[Shared] Attachment ownership check for MCP tools**: MCP tool calls now check whether an attachment belongs to the current conversation, and attachments from other conversations are blocked.
* **\[Shared] Fix for member removal happening at the same time as SSO login**: A removed member's removal is no longer reverted even if they sign in through SSO at the same time.
* **\[Shared] Telegram Bot Token and Text2SQL connection strings stored encrypted**: The Telegram Bot Token and the Text2SQL external database connection URL are now stored encrypted, so plaintext no longer appears in backups.
#### 🛠️ Bug Fixes
* **\[Shared] GPT-5.6 Sol via direct OpenAI stays available**: The GPT-5.6 Sol model connected directly to OpenAI is no longer retired by a scheduled job after deployment, so AI assistants using it are not automatically moved back to the Azure endpoint.
* **\[Shared] Token-by-token streaming for OpenAI models on Bedrock**: AI assistants using GPT-6.1 Sol and GPT-5.6 Sol on Bedrock now show answers as they are generated, instead of waiting for the full response before displaying it all at once.
* **\[Shared] Fallback models are charged by the model that actually answered**: Conversation replies now deduct credits by the model and unit price that actually answered, instead of continuing to use the primary model when a fallback group takes over.
* **\[Shared] Answers blocked by the model provider are no longer charged**: When using platform models, the call whose answer was blocked by the provider's content moderation no longer deducts credits.
* **\[Shared] API documentation aligned with actual behavior and API Key recorded**: Fixed five API documentation descriptions that did not match actual behavior, and API call records now also show the API Key used.
* **\[Shared] Ask AI long tasks no longer fail from expired credentials**: Credentials are now refreshed automatically while Ask AI runs long tasks, so 401 errors no longer appear after 10 minutes, and retries stop after consecutive failures.
* **\[Shared] API key requests with a non-object body now return 400**: When sending a message with an API key, a body that is not a JSON object now returns 400 instead of 500.
* **\[Shared] Files that fail Vision parsing are no longer charged**: If a knowledge base file ultimately fails after Vision parsing, customers are no longer charged for that run.
* **\[Shared] Claude prompt cache markers now sent**: For Claude models on Vertex or direct Anthropic, prompt caching is now actually applied once it is turned on.
* **\[Shared] Text-to-SQL deletion and re-upload fixes**: Deleting Text-to-SQL data files no longer wrongly clears their own tables, and interrupted files can be re-uploaded and completed.
* **\[Shared] Vision Parser resume and billing fixes**: Vision Parser now correctly picks up progress when resuming after an interruption and records usage completely.
* **\[Shared] One-off calls no longer send cache markers**: One-off calls such as memory consolidation and question rewriting are no longer billed at the cache-write price.
* **\[Shared] Real-time information labeled as platform information**: After a tool call, AI assistants no longer mistake the real-time information appended by the platform for something the user said.
* **\[Shared] Credit record fix for turns not billed per message**: In Team, restart-interrupted, and no-charge turns, tool charges are now recorded on that turn's assistant reply.
* **\[Shared] Web search and map tools without a key are no longer charged**: When no key is configured, web search and map tools are now recorded as failed and no longer deduct credits.
* **\[Shared] Multi-item citation matching for the web fetch tool**: The global web fetch tool no longer wrongly reports "not found" for multi-item questions, and fetch failures are recorded as failed calls.
* **\[Shared] Vision parsing rate-limit handling**: When many files use Vision Parser at the same time, concurrency is controlled by quota, and rate-limited pages are automatically queued to continue.
* **\[Shared] Long-term memory no longer answered as a new user message**: When an assistant with long-term memory answers after a tool call, it no longer treats the memory attached by the platform as a new question from the user.
* **\[Shared] OpenAI and Azure cache write credits are calculated at provider prices**: Cache write costs for OpenAI and Azure models are now charged in credits based on the provider's published prices.
* **\[Shared] SSL relaxed domain list applied to image downloads**: The SSL relaxed domain list now also applies to image downloads for text-to-image reference images and image-to-video, so storage using self-signed certificates can be accessed normally.
* **\[Shared] Admin console automatically creates cache write pricing for OpenAI/Azure models**: When an OpenAI or Azure model is added in the admin console, a credit pricing rule for cache writes is now created along with it.
* **\[Shared] On-premises upgrades skip official Agents when a global tool is missing**: If a tool required by an official Agent is missing on the site during an on-premises upgrade, that Agent is skipped and the missing items are listed, instead of the upgrade failing and rolling back.
* **\[Shared] Unauthorized contacts can get MCP tool authorization links**: When tool search is enabled, unauthorized contacts can now receive the authorization link for MCP tools, and the reminder names the services that are not yet authorized.
* **\[Shared] Clear message when registering with an already registered email**: When registering or accepting an invitation with an email that is already in use, "This email has already been registered" is now shown instead of an error.
* **\[Shared] Tool headers with Chinese contact names now work**: API and MCP tools can now be called normally when Chinese contact names are included in headers.
* **\[Shared] Fixed Responses API model reply failures**: For assistants using Responses API models such as GPT-6 Luna, a blank text segment in the tool loop no longer causes the entire reply to fail.
* **\[Shared] API docs now describe fields for invitations and new-member quotas**: The API docs now include request and response formats for unified invitations, default quotas for new members, and quota settings.
* **\[Shared] Reasoning depth adjustable for direct OpenAI GPT-6 Luna/Sol**: Direct OpenAI GPT-6 Luna and Sol now support adjusting reasoning depth on the assistant settings page and in MaiGPT.
* **\[Shared] API key allowlists with only owned models are no longer treated as unrestricted**: When an API key's AI Gateway allowlist lists only the organization's own models, it is no longer mistakenly treated as unrestricted.
* **\[Shared] Models past their retirement date no longer appear in menus**: Platform models past their retirement date no longer appear in model menus and cannot be newly bound, and the admin console adds a retirement status column.
* **\[Shared] AWS Marketplace usage reporting now scheduled**: AWS Marketplace usage is now reported automatically every hour, with retries on failure and a daily reconciliation.
## v2026.10.17
***
### 📌 Highlights
* **\[Shared] Better cache reuse in long AI assistant conversations**: AI assistants using Claude models, such as MaiGPT, previously could rewrite the entire cache when a new question was asked in a long, multi-turn conversation. The system now adds an extra cache anchor at the previous question, so subsequent questions only require processing newly added content. Conversation history reuses the existing cache, reducing repeated work.
### 🗂️ Full Update (by Product Line)
Shared (1 item)
#### 🚀 Core Performance Improvements
* **\[Shared] Better cache reuse in long AI assistant conversations**: AI assistants using Claude models, such as MaiGPT, previously could rewrite the entire cache when a new question was asked in a long, multi-turn conversation. The system now adds an extra cache anchor at the previous question, so subsequent questions only require processing newly added content. Conversation history reuses the existing cache, reducing repeated work.
## v2026.10.16
***
### 📌 Highlights
* **\[Serve] Fixed pagination count timeouts on the conversation history page**: Members who are not organization owners could wait tens of seconds or encounter a timeout when opening the admin conversation history page in organizations with many inboxes. The count now reads only conversations belonging to their organization, reducing the worst-case duration from about 35 seconds to less than 0.1 seconds.
### 🗂️ Full Update (by Product Line)
Serve (1 item)
#### 🛠️ Bug Fixes
* **\[Serve] Fixed pagination count timeouts on the conversation history page**: Members who are not organization owners could wait tens of seconds or encounter a timeout when opening the admin conversation history page in organizations with many inboxes. The count now reads only conversations belonging to their organization, reducing the worst-case duration from about 35 seconds to less than 0.1 seconds.
## v2026.10.15
***
### 📌 Release Highlights
* **\[Work] Personal AI memory switch**: The AI Memory page in your profile now includes a “Let the assistant remember me” switch, configured separately for each organization. Turning it off stops new memories from being saved and existing memories from being used in conversations. Saved memories remain available to view and manage. [User Guide](https://docs.maiagent.ai/work/maigpt/ai-memory)
Personal AI memory switch
* **\[Build] Admin conversations respect assistant tag filters**: When you leave the reference document selection unchanged, admin conversations use the assistant's existing tag filters, matching the default retrieval scope in Web Chat. Explicitly selecting documents still limits retrieval to that selection.
* **\[Shared] Inline citations no longer delay the start of answers**: In standard answer mode with inline citations enabled, answer generation no longer waits for citation data to finish sending. This reduces the wait before an answer starts; citations can follow the initial text. API integrations should allow citation information to arrive after some answer text.
* **\[Work] Workspace file deletion approval**: The file workspace toolbar now shows the number of pending deletion requests. Owners can use version history to review the filename, requester, source conversation, and deadline, then approve or reject each request submitted by an Agent on a member's behalf. [User Guide](https://docs.maiagent.ai/work/file-library)
* **\[Serve] Bulk deletion of Web Chat history**: Web Chat history now has an edit mode for selecting multiple conversations or all currently loaded conversations and deleting them together. If some deletions fail, the result reports success and failure counts and keeps the failed items available for retry. [User Guide](https://docs.maiagent.ai/serve/daily-operations/reply)
### 🗂️ Full Updates by Product Line
Work (24 items)
#### ✨ New Features
* **\[Work] Personal AI memory switch**: The AI Memory page in your profile now includes a “Let the assistant remember me” switch, configured separately for each organization. Turning it off stops new memories from being saved and existing memories from being used in conversations. Saved memories remain available to view and manage. [User Guide](https://docs.maiagent.ai/work/maigpt/ai-memory)
Personal AI memory switch
* **\[Work] Workspace file deletion approval**: The file workspace toolbar now shows the number of pending deletion requests. Owners can use version history to review the filename, requester, source conversation, and deadline, then approve or reject each request submitted by an Agent on a member's behalf. [User Guide](https://docs.maiagent.ai/work/file-library)
#### 😊 User Experience and Interface Improvements
* **\[Work] Live translation models display their providers**: The meeting live translation model menu shows the provider beside each model name. Both remain visible when the menu is collapsed, making the selected model easier to identify.
* **\[Work] Localized file space names and paths**: File library search results and the MaiGPT workspace picker display personal space, organization space, and temporary storage names in the interface language. Custom space and folder names remain unchanged, and the picker guidance now reflects current file usage.
* **\[Work] Generated files appear immediately with links to the library**: AI-generated files appear in the output panel right away, with a processing status while they are saved to the file library. Once archived, the original item provides access to the file's location.
* **\[Work] Reorganized connector and skill menus**: The composer's “+” menu lists only authorized, available connectors. Authorization and adding connectors are handled through “Manage connectors,” and empty skill or tool groups are hidden. This also applies to the classic MaiGPT layout.
* **\[Work] Clearer file deletion and restore messages**: 1. Deletion confirmations identify the file and consequences, and restore conflicts show “Restore failed” with a reason; 2. Deeply nested folder names are truncated to leave room for actions; 3. Shared spaces now include relevant guidance and empty states to clarify which actions are available.
* **\[Work] Organization sharing list pagination and layout updates**: The organization sharing page separates the data area from the pagination bar at the bottom, with 10, 20, 50, or 100 items per page. The title and actions stay in place while you browse longer sharing histories.
* **\[Work] File library actions stay in place**: The file library pins the title, search, upload, path, and toolbar above the data area. The file list scrolls independently so you can keep managing files from the same position.
* **\[Work] Meeting list pagination and horizontal browsing updates**: The meeting records list keeps its header and filters pinned. It defaults to 20 items per page, with options for 10, 20, 50, or 100. A floating horizontal scrollbar and a compact action menu make wide tables easier to browse.
* **\[Work] Consistent list menus and confirmations**: List action menus adjust their opening position to fit the available space, and deletion uses consistent confirmations and result messages. File renames can be undone from the action notification, with an explicit message when undo is no longer available.
* **\[Work] Presentation shortcut and source selection updates**: The presentation shortcut in classic MaiGPT now uses a fixed length of 10 slides, and the panel closes after confirmation. Sources supplied by templates appear in the composer alongside skills and tools, grouped by count when space is limited.
#### 🛠️ Bug Fixes
* **\[Work] Folders follow list and thumbnail view changes**: In the file library's list view, folders appear at the top of the same table as files, and clicking a folder name opens it. Thumbnail view keeps folder cards, and Select All still selects only files.
* **\[Work] Attachment expiry countdown displays correctly**: The conversation attachment list correctly shows the remaining retention days and marks expired attachments, helping you handle files before their retention period ends.
* **\[Work] Full model names remain visible**: The MaiGPT model menu adjusts its width and narrow-screen wrapping to prioritize complete model names. Hover over a model to see its full name and provider.
* **\[Work] Clear loadable status for attachments**: Attachments that need no parsing consistently show “Ready to load” in file library lists, search results, and details. This covers archived attachments, Agent output, and restored files, making it easier to tell whether AI can use them.
* **\[Work] Mobile file library actions are no longer clipped**: On mobile, the file library places search and action buttons on separate rows and provides a “Spaces and folders” drawer for switching locations. Details, guidance, and version history drawers adapt to the screen width.
* **\[Work] Shared pages load their styles correctly**: When shared pages open from the classic layout, newer sections also load their full styles and organization colors, preventing unstyled lists or dialogs.
* **\[Work] Transcript content survives refresh failures**: If refreshing the same meeting transcript fails, the content already displayed stays visible with a failure message. Switching transcript versions also rechecks the source to avoid showing content from a different version.
* **\[Work] Connector activation notifications open management directly**: Links in connector activation notifications open MaiGPT or Work and show “Manage my connectors.” Notifications for another organization first switch to that organization. Existing notification links also work.
* **\[Work] Knowledge base search preserves selections**: 1. Searching after selecting knowledge bases leaves previously unselected sources unchecked; 2. Links from the knowledge base page to MaiGPT correctly expand the panel; 3. Switching conversation platforms reloads display settings to avoid carrying over knowledge cards from the previous platform.
* **\[Work] Cancelled recordings no longer reopen the microphone**: If a recording is stopped while connecting, ended by the server, or replaced by a new meeting, the old connection cleans itself up when it completes. It no longer opens the microphone, starts backup recording, or stops a subsequently created meeting.
* **\[Work] Switching Agents no longer sends messages to the old conversation**: Switching Agents first clears the previous conversation's send target. If the new conversation fails to load, messages cannot be sent to the earlier conversation, keeping the displayed name consistent with the assistant that actually replies.
* **\[Work] Ask AI history loads with the member's identity**: Admin Ask AI history requests now include the current member's identity so authorized conversations can load, and update authentication information as the login state changes.
Serve (12 items)
#### ✨ New features
* **\[Serve] Bulk deletion of Web Chat history**: An edit mode in Web Chat history lets visitors select multiple conversations or all currently loaded conversations for bulk deletion. If some deletions fail, the results show success and failure counts and retain failed items for retry. [User guide](https://docs.maiagent.ai/serve/daily-operations/reply)
#### 😊 User experience and interface improvements
* **\[Serve] Service impact warning before switching Copilot modes**: Selecting Copilot's “Every incoming message” mode now explicitly warns that AI will not reply automatically, automatically hand off to a human agent, or speak during web calls, and reminds administrators to ensure staff can respond promptly.
* **\[Serve] Consistent Traditional Chinese terminology and localized Web Chat prompts**: Web Chat now uses common Taiwanese terminology in Traditional Chinese, with consistent wording for branched conversations. Add-to-home-screen, installation, and connection error prompts follow the current language.
* **\[Serve] Warning before sending external channel attachments as links**: When an attachment type cannot be sent directly through LINE or Instagram, a prompt before sending explains that a download link will be sent instead, helping staff confirm what the customer will receive.
* **\[Serve] Return conversations to AI from any tab**: The customer service conversation action “Return to AI Agent” is available on every tab. Supervisors and colleagues no longer need to assign a conversation to themselves and switch to “Mine” first. On narrow screens, the status bar wraps to keep the button visible.
#### 🔐 Security and governance enhancements
* **\[Serve] Interactive cards restrict external connections by default**: Web Chat MCP interactive cards block external website connections when no allowed origins are configured. Administrators must explicitly allow origins for services that need external access.
* **\[Serve] Permission checks before adding channels**: Members without the “Allow adding conversation platforms” permission see a disabled channel connection entry with an explanation. Direct access to the creation page is also blocked, so they do not complete a form only to discover they cannot save it.
#### 🛠️ Bug fixes
* **\[Serve] Authorization cards no longer reuse another contact's status**: Switching contacts or clearing the identity in Web Chat first clears the previous contact's connection status. If loading fails, the authorization entry for the current identity appears instead of incorrectly showing another contact's connected status.
* **\[Serve] Complete reply endings**: Once Web Chat receives the saved reply, it fills in any trailing text missed during streaming and verifies the order, keeping the live answer consistent with the answer shown after refresh.
* **\[Serve] Customer service conversations no longer go blank when switching tabs**: 1. Switching tabs while messages are loading now updates the conversation correctly instead of leaving it blank; 2. List previews strip Markdown symbols, and completion messages correctly refer to conversations or messages; 3. Copilot manual mode instructions match the actual entry point.
* **\[Serve] Human agent waiting status syncs after reconnection**: When Web Chat replays missed events after reconnecting, it also processes cancellations of the current handoff, staff takeover, and returns to AI. Stale waiting states are cleared and the input is unlocked without waiting for the queue to time out.
* **\[Serve] Reloading settings preserves queue-period drafts**: Reloading human handoff settings preserves unsaved queue-period changes, including text entered while a save is pending. Unchanged fields still update to the latest server values.
Build (18 items)
#### 😊 User experience and interface improvements
* **\[Build] Clearer schedule actions and execution statuses**: Schedule toggle tooltips now explain the upcoming action explicitly. Execution details and filters consistently display “Partially successful” so incomplete runs are easier to identify.
* **\[Build] Hook logs identify inspection sources**: Hook execution logs, filters, and details now use readable labels for trigger points, actions, and match results. A new inspected-content field distinguishes user messages, knowledge base retrieval results, and other sources.
* **\[Build] Feedback counts in satisfaction charts**: Satisfaction chart tooltips show the percentage alongside like and dislike counts for hourly, daily, and monthly statistics, helping distinguish a lack of feedback from actual dissatisfaction.
* **\[Build] Activation guidance for disabled databases**: The AI assistant's database selector now lists disabled databases as unavailable and directs members to activate them on the database page. Linked but disabled databases and their settings pages also explain that assistants cannot currently query them.
* **\[Build] New assistants truncate long tool results by default**: Tool result truncation is enabled by default for AI assistants created in the admin console, matching the platform default. It can still be adjusted in advanced settings as needed.
* **\[Build] Agent marketplace remembers the last tab**: The Agent marketplace remembers the last tab used in the same browser and returns to it when reopened or refreshed. Clearing cookies or switching devices or browsers restores the default tab.
* **\[Build] Skill availability notice outside Agent mode**: When an assistant uses a response mode other than Agent mode, skill settings explain that skills are available only in Agent mode and provide a link to response mode settings, preventing inactive skills from being mistaken for a channel issue.
* **\[Build] FAQ imports report added and skipped counts**: The FAQ bulk upload button now says “Upload and add,” explaining that rows with duplicate questions are skipped without overwriting answers. Completion notifications and import logs separately report added entries, skipped duplicates, and errors.
* **\[Build] Centralized marketplace search and management**: The Agent marketplace places current tab information, management access, and search in the title bar. Category and favorites filters stay above, while only the Agent card area scrolls.
#### 🛠️ Bug fixes
* **\[Build] Admin conversations respect assistant tag filters**: Unless reference document selections have been changed, admin conversations use the assistant's existing tag filters, matching Web Chat's default retrieval scope. Explicitly selecting documents still limits retrieval to that selection.
* **\[Build] Confirm authorized roles before creating skills and tools**: When creating skills, tools, or Flex Message interface tools, saving redirects members to the permissions tab for confirmation if they have not reviewed it yet, preventing newly created resources from being visible only to their owner.
* **\[Build] Specific reasons for tool save failures**: Rejected saves for tools and interactive interface tools now show the actual reason, such as a duplicate name, and avoid duplicate error dialogs so members can correct fields directly.
* **\[Build] Agent previews play with reduced motion enabled**: Agent marketplace previews still play demo conversations and walkthroughs message by message when reduced motion is enabled, and can be replayed, while omitting typewriter animations and smooth scrolling.
* **\[Build] Correct status labels for crawl failures**: Failed web crawls now appear as Failed across the task list, details, page list, and schedule history, matching the filters. Specific failure reasons remain available.
* **\[Build] Evaluation reports identify the correct assistant**: The AI Assistant field in detailed reports for individual evaluation questions now shows the assistant actually evaluated, matching the evaluation settings. It stays empty until the data loads to avoid confusion.
* **\[Build] AI search results and summaries use the same query**: Knowledge base AI search uses the text submitted at that moment for results, summaries, and search history. Submitting different keywords during a search starts a new query, and older results that arrive later no longer overwrite the new search.
* **\[Build] Assistant settings recognize your own updates**: After adding or removing knowledge bases, tools, connectors, skills, or databases in assistant settings, you can save the name and instructions without your own immediate saves being mistaken for another user's updates. Actual concurrent edits are still checked.
* **\[Build] Tool editing discards stale responses**: Closing the tool editor or switching tools now discards late content, authorization status, and tool lists from the previous session, preventing settings for tool A from being saved to tool B.
Shared (26 items)
#### 🚀 Core Performance Improvements
* **\[Shared] Inline citations no longer delay the start of answers**: In standard answer mode with inline citations enabled, answer generation no longer waits for citation data delivery, reducing the initial wait. Citations may be added after text starts arriving. API integrations should allow citation information to arrive after some of the text.
#### 😊 User Experience and Interface Improvements
* **\[Shared] Consistent announcement reminders and read status**: 1. Feature announcements appear only on login, loading, or page navigation; closing one no longer immediately opens the next. 2. Organization announcements follow the configured reminder mode, and brief notifications start their countdown only after the dialog closes. 3. Clicking an announcement link marks it as read, preventing another reminder after a refresh.
* **\[Shared] Admin navigation and page names match their destinations**: 1. Login source descriptions and onboarding invitation templates now reflect the actual operations. 2. The call center directs members to an available page instead of always opening number settings. 3. Credit export and conversation platform tab titles match their content.
* **\[Shared] Localized real-time voice call billing descriptions**: The billing method for real-time voice call duration now includes names and descriptions in each interface language, helping members understand how call time counts toward usage.
* **\[Shared] Accurate email invitation validity information**: Email invitations now explain that they remain valid until the recipient joins or the invitation is revoked, without displaying an expiration period. Invitation links still support an expiration date, and switching invitation methods preserves the selected period.
* **\[Shared] Organization settings explain what members can edit**: Assigning AI assistants now explains that their conversation platforms are assigned as well. Announcements and model invocation permissions are read-only for non-owners, with explanations, reducing permission errors after members attempt an operation.
* **\[Shared] Correct API key error response guidance**: The API key management drawer now explains that subsequent calls with a disabled key return 401, while calls to models that have not been enabled return 404, helping developers troubleshoot integrations using the actual responses.
* **\[Shared] Visible default resource permission assignments**: Permission settings for new resources now highlight the applied default option and summarize role assignments after creation. Non-owners also see an explanation of the scope they can assign, avoiding the impression that the entire organization has been granted access.
* **\[Shared] SSO confirmation explains that existing passwords remain usable**: The SSO activation dialog clearly states that members who have set a password can still sign in with their username and password, helping administrators plan account management around the actual login rules.
* **\[Shared] Settings and integration descriptions match actual behavior**: 1. Message Webhooks offer only POST, PUT, and PATCH, and explain the direction of visitor message and assistant reply delivery. 2. Delivery logs correctly show the conversation ID, and skill package re-upload is available only for skills created by upload. 3. Knowledge base similarity threshold guidance matches the actual filtering behavior.
* **\[Shared] Integration examples match the actual interfaces**: Admin Source ID examples now wait for the SDK to load before initializing identity. API instructions use the correct Api-Key header and /api/v1/ paths, and file upload examples include the required steps.
#### 🔐 Security and Governance Enhancements
* **\[Shared] Isolated canvas previews**: The admin console and Web Chat strengthen isolation for generated canvases, restricting access to login information and actions on the outer page. Interactive previews and PDF export remain available, and unsafe content is sanitized before display.
* **\[Shared] Cloud analytics removed from on-premises builds**: New on-premises admin console and Web Chat builds remove cloud analytics tracking settings to prevent usage events from being sent to the platform's cloud analytics service. This applies after deploying the new on-premises images.
* **\[Shared] Navigation restrictions for unavailable pages**: Internal demo pages are no longer accessible in production. When the billing management entry is disabled, opening its URL directly redirects to an available organization page, preventing use of unreleased features.
* **\[Shared] Real-time connection URLs no longer carry login credentials**: Real-time connections in the admin console and Web Chat now pass login credentials through authentication data, keeping them out of connection URLs and access logs. Logging out clears existing connections, and reconnections use the current login identity.
#### 🛠️ Bug Fixes
* **\[Shared] Consistent canvas source download formats**: Mermaid chart downloads now use .mmd across all entry points. Map canvases offer only coordinate source files in .json, and downloading map source files from the admin console works again.
* **\[Shared] Complete contact details in on-premises onboarding**: MaiAgent-branded on-premises sites provide a default contact email when no support email is configured. Custom brands without an email hide the empty contact entry to avoid opening an email without a recipient.
* **\[Shared] Canvas exports match the content at download time**: Canvas exports check the content and preview version captured when you click download. Switching canvases or recompiling while waiting cancels the export to prevent downloading the wrong version. When exporting a PDF from code mode in Web Chat, the preview for that version is completed first.
* **\[Shared] Currency symbols no longer trigger math formatting**: NT$, US$, HK$, and other monetary amounts in replies, citations, and document previews retain their original formatting. Multiple amounts in the same paragraph no longer turn into italics or math loading indicators.
* **\[Shared] On-premises documentation links open the correct help pages**: Model selection recommendations and role instruction help in the on-premises admin console link to their corresponding documentation pages. Default help links also work when no documentation site is configured, preventing invalid URLs.
* **\[Shared] Multiple cards in one reply appear immediately**: When a reply generates multiple interactive cards or sets of quick replies, the admin console and Web Chat retain each one in order without overwriting earlier cards. The live display now matches what appears after a refresh.
* **\[Shared] Model tool-calling capabilities are labeled correctly**: The model menu uses actual capabilities from the model catalog to display the function-calling icon and description, making models that support tool calling easy to identify.
* **\[Shared] Identity synchronization permissions and credit alerts are aligned**: Dedicated permissions now control access to the identity synchronization page, and missing permission names have been added. Low-balance thresholds on personal and role credit cards follow the organization's alert settings, keeping cards consistent with notifications.
* **\[Shared] File-writing progress and errors display correctly**: When an assistant calls other tools while writing a file, Web Chat file cards no longer get stuck blank and the admin console no longer exposes file contents. Errors during generation appear as regular conversation text instead of being hidden inside the file.
* **\[Shared] Links to newly joined organizations open directly**: When an organization notification or Email link points to an organization missing from the cached list, the system refreshes membership before checking access. Members can enter newly joined organizations without refreshing the entire page first.
* **\[Shared] Credit details use the latest filter results**: When Credit usage history filters change rapidly, the list accepts only results from the latest query to avoid mixing transactions from other categories. Loading indicators also reflect only requests for that list.
## v2026.10.12
***
### 📌 Release highlights
* **\[Work] Outlook email file (.eml/.msg) parsing**: Knowledge bases and conversation attachments can now process Outlook email files directly: 1. After you upload an .eml or .msg file, the system parses the message body and sender information and lists its attachments; 2. It also extracts content from embedded document attachments, such as quotes, PDFs, and Excel files, so AI can answer questions about their details; 3. Email attachments forwarded through Outlook are recursively expanded; 4. S/MIME-signed or encrypted emails are handled according to their status, and encrypted content that cannot be read is marked; 5. Deleting an email file also removes the attachments extracted from it; 6. TNEF (winmail.dat) files are detected and marked as present without being expanded. Useful for teams organizing customer emails and quote exchanges in a knowledge base or dropping them into a conversation to ask questions. [User guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Serve] Automatic return to AI when human support does not reply in time**: Conversation platforms now offer a timeout setting for returning conversations to AI. After a human agent takes over, if no one replies to a visitor's message within the number of seconds you specify, the system automatically hands the conversation back to the AI assistant and can send a custom notification. The timer starts when the agent first takes over. Scheduling failures no longer drop handoff notifications, and messages arriving at the moment of takeover are captured. Useful for teams with limited support staff who want to avoid leaving visitors waiting. [User guide](https://docs.maiagent.ai/build/conversation-settings/human-handoff)
* **\[Shared] Bulk role permission settings and effective member permissions**: Organization management adds two permission governance capabilities: 1. Apply permission settings to multiple MaiAgent roles at once instead of editing them individually; 2. View a member's effective permissions and see which role grants each one. Permission screens have also been refined: role permission names are now all in Chinese and match the UI, role permissions are sorted by display name, missing translations for Hook and conversation platform analytics permissions have been added, and the quota self-check screen now includes the alert threshold. [User guide](https://docs.maiagent.ai/org/roles)
* **\[Work] Work Agent file workspaces: mount File Library folders on custom and official Agents**: File Library folders can be mounted on Work Agents as workspaces: 1. Custom Agents can have a bound workspace and can allow users to mount their own; 2. Official Agents allow users to mount their own folders; 3. Mounting now supports any nested subfolder, and folder file counts include subfolders. Files modified by an Agent in a workspace can be saved back to File Library. [User guide](https://docs.maiagent.ai/work/file-library)
* **\[Work] Bulk deletion of your conversations**: Select and delete multiple conversations you own from the conversation list in one action, making it faster to clean up a large backlog. [User guide](https://docs.maiagent.ai/work/maigpt)
* **\[Build] OpenAI-compatible speech endpoints in AI Gateway**: AI Gateway adds OpenAI-compatible speech-to-text and text-to-speech endpoints. Existing applications using OpenAI speech APIs can connect by changing only the URL and API key, with Credit deducted according to actual usage. [User guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Serve] Pass quoted message content to AI**: On platforms that support quoted replies, such as LINE, when customers quote an older message to ask a question, the system sends the quoted content to AI as well, so the answer addresses the specific message. [User guide](https://docs.maiagent.ai/build/choose-channel)
* **\[Shared] Signature verification and automatic retries for message Webhooks**: Message event notifications (Webhooks) now include signatures and timestamps, allowing receivers to verify that notifications come from MaiAgent and guard against replay attacks. The system automatically retries delivery after temporary failures. [User guide](https://docs.maiagent.ai/org/developer)
### 🗂️ Full updates (by product line)
Work (18 items)
#### ✨ New features
* **\[Work] Outlook email file (.eml/.msg) parsing**: Knowledge bases and conversation attachments can now process Outlook email files directly: 1. After you upload an .eml or .msg file, the system parses the message body and sender information and lists its attachments; 2. It also extracts content from embedded document attachments, such as quotes, PDFs, and Excel files, so AI can answer questions about their details; 3. Email attachments forwarded through Outlook are recursively expanded; 4. S/MIME-signed or encrypted emails are handled according to their status, and encrypted content that cannot be read is marked; 5. Deleting an email file also removes the attachments extracted from it; 6. TNEF (winmail.dat) files are detected and marked as present without being expanded. Useful for teams organizing customer emails and quote exchanges in a knowledge base or dropping them into a conversation to ask questions. [User guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Work] Work Agent file workspaces: mount File Library folders on custom and official Agents**: File Library folders can be mounted on Work Agents as workspaces: 1. Custom Agents can have a bound workspace and can allow users to mount their own; 2. Official Agents allow users to mount their own folders; 3. Mounting now supports any nested subfolder, and folder file counts include subfolders. Files modified by an Agent in a workspace can be saved back to File Library. [User guide](https://docs.maiagent.ai/work/file-library)
* **\[Work] Bulk deletion of your conversations**: Select and delete multiple conversations you own from the conversation list in one action, making it faster to clean up a large backlog. [User guide](https://docs.maiagent.ai/work/maigpt)
* **\[Work] File Library storage usage and a dedicated Credit category**: Two additions make File Library usage more transparent: 1. Organization overview can report File Library storage usage; 2. Credit usage records include a dedicated service category for File Library, with derived content storage accounted for separately by source, making storage and Credit spending easier to track. [User guide](https://docs.maiagent.ai/work/file-library)
* **\[Work] Meeting context for speech recognition in meeting records**: Speech recognition in meeting records now uses meeting context: 1. Calendar invitation content is applied before a meeting is created, providing recognition cues such as the meeting room name and participants; 2. Context is built and sent to the Agent before live recognition starts; 3. Offline recognition also receives context when importing files or retranscribing; 4. A feature flag lets administrators control the capability. This improves recognition of specialized terms and people's names. [User guide](https://docs.maiagent.ai/work/meeting-records)
* **\[Work] Separate AI memory pricing and transparent usage**: AI memory Credit usage is now clearer: 1. Memory extraction is billed separately against the quota of the member being remembered; 2. Credit usage records include an AI memory service category; 3. Response input details record the number of tokens used by AI memory; 4. Summaries filter out personal information, and malformed output is not charged. [User guide](https://docs.maiagent.ai/work/maigpt)
* **\[Work] Structured organization of AI memory**: AI memory is now better organized: 1. Memories are stored by category, and new information merges with existing memories to avoid duplicates; 2. User facts are extracted together after a conversation has been quiet for a while; 3. Memory topics appear as titles in the user's language for all organizations. [User guide](https://docs.maiagent.ai/work/maigpt)
#### 🚀 Core performance improvements
* **\[Work] Faster, less expensive memory background tasks**: Memory background operations use fewer resources: 1. Asking the assistant to remember something no longer requires an extra round trip to the main model, and memory included in each turn is capped at 600 tokens; 2. Background calls use system task models without thinking, follow organization rules, and fall back to the main model only on failure; 3. Long-conversation consolidation uses system task models, with cost guardrails configurable in Admin; 4. If a system model fails, the task retries once using the next system task model; 5. A model response stating that there are no facts no longer triggers a fallback rerun, reducing excessive fallback rates.
* **\[Work] Higher prompt cache hit rates for long conversations**: Prompt caching for OpenAI, Azure, and Claude is more stable: 1. Long-term memory is moved out of the cached prefix so Claude no longer rewrites the entire cache on every conversation turn; 2. Short-term memory compression now reduces usage to 60% of the budget and compresses in batches when the protected portion itself is too large, keeping the start of history stable across turns; 3. The current time is moved to the end of the message, and each user message carries its send time, allowing long conversation histories to hit the cache; 4. Cache routing now uses the assistant as the key, allowing new conversations to hit the cache too (GPT-5.6 previously missed every time).
* **\[Work] More concise Canvas Agent prompts**: Tool results are no longer repeated in the Canvas Agent prompt after handoff, shortening prompts and reducing model costs.
#### 🔐 Security and governance enhancements
* **\[Work] GPT image generation and editing no longer retain content at OpenAI**: Content is no longer stored at OpenAI when generating or editing images with GPT. Edits now send the locally stored previous image, reducing the risk of image data remaining with a third party.
#### 🛠️ Bug fixes
* **\[Work] AI memory truncation and toggle issues fixed**: 1. When older tool results are truncated, knowledge base content retains 100 characters and any IDs removed by truncation; only the copy written to history is truncated; 2. IDs are also preserved when JSON printed by Code Interpreter is truncated; 3. When AI memory is off, the assistant no longer promises to remember things; 4. Fixed a regression from these changes: with memory off, the assistant no longer responds to ordinary requests with “I cannot assist”; 5. When a message is recalled during a response, memory still retains the attachment IDs of photos from the same turn.
* **\[Work] File Library downloads and conversation attachment issues fixed**: 1. Downloads of File Library conversation attachments and generated files no longer return a 500 error when the file exists; 2. Downloads retain the display filename instead of using a 32-character hash; 3. Regular members can open File Library conversation attachments without loading failures (listing, downloading, and early deletion all work); 4. Lists and downloads of conversation-generated files no longer return 500 errors after the assistant is deleted; 5. MaiGPT conversation attachment lists include the workspace source (Workspace › Folder).
* **\[Work] Fixed file library version history and search**: 1. Version history messages follow the interface language; 2. Search results include creation time and the last updater; 3. File counts can include subfolders; 4. Version history identifies who approved Agent deletions, renames, moves, and edits, along with the source conversation. Source conversations for generated files show their display names.
* **\[Work] Fixed confirmation and write-back for Agent file operations**: 1. Agent file edits and renames succeed on the first confirmation; 2. Office extraction results explicitly instruct the Agent to write changes back to the file library; 3. When the Agent modifies a sandbox copy of a workspace file, it receives a reminder to write it back; 4. Files generated by code are saved to the file library's temporary storage as soon as they are written, without opening the Output panel; 5. Generated images and videos are recorded directly without queueing.
* **\[Work] Fixed file library images, folders, and notifications**: 1. Uploaded file library images are immediately marked as ready to load instead of showing a knowledge base parsing failure; 2. Empty folders and file workspaces can be deleted; 3. Deletion approval notifications include a workspace link; 4. The Output panel in Work conversations reads from the file library and identifies the owner from the source when the contact's user is empty; 5. The current-conversation file tool omits image lists for non-Agent queries, all tools block removed files, and Canvas lists only the latest version.
* **\[Work] Fixed Canvas and presentation assistant responses**: 1. After calling a nonexistent tool, the Canvas agent is prompted to write directly to Canvas instead of switching to location search; 2. A route map no longer appears as raw JSON after a map was drawn in the previous turn; 3. AI presentation responses include the current turn's image and layout status, so unconfirmed images are no longer described as already inserted into the presentation.
* **\[Work] Fixed voice and meeting record issues**: The meeting transcription context field now has a database default, preventing meeting creation failures during rolling deployments. Meetings now start with a context prepared in advance.
Serve (15 items)
#### ✨ New Features
* **\[Serve] Automatically return conversations to AI when human support does not respond in time**: Conversation platform settings let you specify how many seconds visitors can wait for a reply after a human agent takes over. If no one responds within that period, the system returns the conversation to the AI assistant and can send a custom notice. The timer starts at the initial takeover; scheduling failures no longer suppress handoff notifications, and messages arriving at the moment of takeover are not missed. This helps teams with limited staffing avoid leaving visitors waiting without a response. [User Guide](https://docs.maiagent.ai/build/conversation-settings/human-handoff)
* **\[Serve] Provide quoted message content to AI**: On LINE and other platforms that support quoted replies, the system includes the referenced content when a customer asks about an earlier message, helping the AI answer the intended question. [User Guide](https://docs.maiagent.ai/build/choose-channel)
* **\[Serve] Reassign returning customers to the original support agent after resolution**: An option in conversation platform settings prioritizes the previous support agent when a customer sends another message within a specified period after the conversation was marked resolved. Customers can continue without explaining the background again. The option is off by default and can be enabled when needed. [User Guide](https://docs.maiagent.ai/build/conversation-settings/human-handoff)
* **\[Serve] Automatically configure and verify LINE Webhooks**: Creating a LINE conversation platform now sets the Webhook URL and verifies connectivity for you, removing the manual step of pasting the URL into the LINE console. [User Guide](https://docs.maiagent.ai/build/choose-channel)
* **\[Serve] Detect other subscribed Apps and conversation routing status on Messenger and Instagram**: When a Messenger or Instagram page has other subscribed Apps, the system detects them and recognizes the conversation routing standby state to avoid competing replies. [User Guide](https://docs.maiagent.ai/build/choose-channel)
* **\[Serve] LINE group speaker lists and member counts**: A new API retrieves the list of members who have spoken in a LINE group conversation and the total group membership, helping support teams and external systems understand participation. [User Guide](https://docs.maiagent.ai/build/choose-channel)
* **\[Serve] Send support videos and voice messages in LINE's native formats**: Videos and voice messages sent by human agents through the conversation platform can now be played directly inside LINE instead of arriving as links. [User Guide](https://docs.maiagent.ai/build/choose-channel)
* **\[Serve] Agent can process Telegram voice messages, videos, and audio files**: The AI assistant can now transcribe and understand Telegram voice messages and videos, including audio and video sent as files. For videos, it extracts the audio track for transcription and reports the video length. [User Guide](https://docs.maiagent.ai/build/choose-channel)
* **\[Serve] Agent can see group messages and speakers without an @ mention**: In LINE, Telegram, Slack, Teams, and LINE WORKS groups, the Agent now receives messages that do not mention it, along with speaker information, so it can understand the full discussion before responding. [User Guide](https://docs.maiagent.ai/build/choose-channel)
* **\[Serve] Include conversations without questions in Web Chat contact history**: The history list for logged-in Web Chat contacts now also shows conversations with no messages, allowing Ask AI to reuse a conversation before a question has been asked. [User Guide](https://docs.maiagent.ai/serve/daily-operations/reply)
* **\[Serve] Per-second voice billing for OpenAI gpt-live-1**: Voice support now separates charges for this model's voice layer, measured in seconds, from backend reasoning tokens. Credit consumption records more closely reflect actual costs. [User Guide](https://docs.maiagent.ai/serve/voice)
* **\[Serve] Voice monitoring: delegation links, summaries, and dashboard metrics**: 1. Sessions show their delegation relationships; 2. The session list includes a summary field with filtering and sorting; 3. The dashboard includes latency and delegation metrics; 4. Historical sessions and legacy Agents also have conversation turn counts; 5. Filters support selecting multiple assistants. [User Guide](https://docs.maiagent.ai/serve/voice)
#### 😊 User Experience and Interface Improvements
* **\[Serve] Present LINE stickers to AI as text or images**: 1. LINE stickers are saved as "\[Sticker: keywords]", and attachment filenames no longer contain sticker IDs; 2. Without keywords, the model receives the sticker image; with keywords, it still receives only text; 3. Deep Research's internal assistants and image limits no longer mistakenly include LINE stickers.
#### 🛠️ Bug Fixes
* **\[Serve] Fixed LINE, Messenger, Instagram, Email, and WhatsApp channel issues**: 1. Filled gaps in sending and receiving human support attachments on LINE, Messenger, and Instagram; 2. Documents received by Email become searchable conversation files, as on other platforms; 3. WhatsApp attachments uploaded from the support console are uploaded to Meta before sending, and AI image replies send only the image without the raw markdown; 4. Messages unsent on LINE or Instagram or deleted on Slack no longer leave their original text accessible to AI; 5. LINE group mentions use name boundaries to prevent incorrect replies to names sharing the same prefix; 6. Resending the same message on LINE or Team+ no longer creates false failure records or alerts; 7. LINE monthly message quotas and favorite conversation platforms no longer return 500 errors when non-owner members belong to multiple groups; 8. Anonymous visitor conversation data now includes the conversation platform ID.
* **\[Serve] Fixed conversation processing status and Copilot suggestions**: 1. Marking a conversation as read no longer overwrites the results of switching back to AI Agent or assigning the conversation with old values; 2. Conversations outside AI auto-reply hours (including when all seven days are disabled) no longer report their progress as being processed by AI; 3. Support Copilot generates suggestions using the conversation as it stood at the time of that visitor message, and drafts are no longer written back to AI memory.
Build (17 items)
#### ✨ New Features
* **\[Build] AI Gateway adds OpenAI-compatible audio endpoints**: Applications already using OpenAI audio interfaces can connect by changing just the URL and key, with new endpoints for speech-to-text and text-to-speech. Credits are deducted based on actual usage. [User Guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Build] Filter the AI assistant list by creator**: Organizations with many assistants can quickly find those created by a specific member, making inventory reviews and handovers easier. [User Guide](https://docs.maiagent.ai/build/agent-builder)
* **\[Build] New assistants default to internal conversations only, with configurable Web Widget/API integration defaults**: To reduce the risk of unintentionally exposing external channels: 1. Newly created AI assistants enable only internal conversations by default; Web Widget and API integration can be enabled separately during creation; 2. Organization settings let administrators centrally define whether conversation platforms and API integration are enabled by default for new assistants. [User Guide](https://docs.maiagent.ai/build/agent-builder)
* **\[Build] Routing entry node added to the team canvas**: At the start of a team workflow, incoming messages can be evaluated and assigned to the appropriate member Agent, making routing logic clearer for multi-Agent teams. [User Guide](https://docs.maiagent.ai/build/agent-builder)
* **\[Build] Filter conversation records by tool and export tool usage**: AgentOps provides more complete tracing: 1. Conversation records can be filtered by the tools used, with tool usage available for export; 2. Built-in tool calls appear in tool execution records; 3. Step cards in team execution records show interaction types and input/output summaries, making workflows easier to investigate. [User Guide](https://docs.maiagent.ai/build/agent-ops)
* **\[Build] MCP Apps: interactive card support**: Web channel conversations now advertise interactive card capabilities to MCP servers, allowing servers built with the official ext-apps SDK to display cards directly in conversations. [User Guide](https://docs.maiagent.ai/build/tools)
* **\[Build] Estimate Credit costs before dataset evaluations**: AgentOps provides a pre-run estimate of required Credits, also available through an API, to help avoid budget overruns on large evaluations. [User Guide](https://docs.maiagent.ai/build/agent-ops)
#### 🚀 Core Performance Improvements
* **\[Build] Lower document parsing costs: Vision Parser resumption and reuse**: Large documents cost less to parse: 1. When output is truncated, complete pages are preserved and only the remaining pages are resubmitted; 2. Page-level checkpoints let interrupted runs resume with unfinished pages only; 3. Reuploading files with identical content reuses existing parsing results without resubmitting to Bedrock; cache connections have timeouts so parsing continues even if Redis hangs; 4. Vision Parser can use Azure OpenAI (GPT-6.1 Sol).
* **\[Build] MCP tool list caching**: Tool lists (tools/list) are cached per identity and sorted consistently. Repeated loads no longer request the list from the MCP server every time, and catalog ordering is more stable.
* **\[Build] Lower evaluation judge call costs**: Dataset evaluations acquire wallet locks later during Credit deduction, reducing peak queueing for parallel runs. Judge calls also stop writing to the prompt cache.
#### 😊 User Experience and Interface Improvements
* **\[Build] Clearer model selection and lists**: 1. All model selectors in the admin interface distinguish models with identical names; 2. Cards for eight official retail Agents in the Agent marketplace now have specific descriptions; 3. The automatically populated default configuration for STT + LLM + TTS voice setups uses Taiwanese Mandarin instead of English; 4. Call titles use the first utterance, and replies spoken by the assistant are labeled as voice transcripts.
* **\[Build] Clearer reasons for MCP connection errors**: When configuring MCP tools: 1. Failed tool retrieval reports the cause (credentials, 401, DNS, connection, or timeout) instead of always saying no tools are available; 2. Server rejections (401, blocked redirects, 5xx) are reported immediately, instead of showing a timeout after 30 seconds; 3. Error events no longer send URLs and credentials to monitoring; 4. API tool Query parameters mask only values resembling credentials, passing ordinary data to AI unchanged.
#### 🔐 Security and Governance Enhancements
* **\[Build] Tighter permissions for Hooks, teams, slides, and templates**: 1. Only owners or creators can modify or delete Hooks and teams; Hook execution records contain only visible AI assistants; 2. Public AI slide sharing links can be disabled, with creation and deactivation restricted to the creator or organization owner; the AI editing entry point is displayed according to permissions; 3. Assistants created from templates are no longer automatically shared with the default role (the whole company); 4. Text to SQL in the shared uploaded-file library uses a table allowlist to block system catalogs and dangerous functions; 5. Test dataset permissions cover all reads, and test question generation can use only knowledge bases readable by the member.
#### 🛠️ Bug Fixes
* **\[Build] Fixed AI assistant tools, skills, and image understanding**: 1. If a file write is left unfinished, the model is prompted to finish and continue; 2. Rewriting the same file in one turn is no longer blocked as a duplicate call; 3. With tool search enabled, bound knowledge base retrieval tools load directly, and mandatory reference instructions are added only when knowledge base content is retrieved; 4. When retrieval results exceed the tool output limit, complete clauses are retained by ranking and only low-scoring nodes are dropped; 5. Attachment parsing tool descriptions recommend only parsers enabled in the current environment; 6. Models receive images when tools revisit them and no longer invent receipt content; 7. If generation fails after a video confirmation card is submitted, the failure message appears below the video task submission reply; 8. Word-to-PDF conversion preserves DFKai-SB and PMingLiU fonts, and charts specifying Windows fonts no longer show boxes in place of Chinese characters.
* **\[Build] Document parsing (OCR) fixes**: 1. The OCR parser now applies PDF page rotation, so documents scanned horizontally and then displayed rotated no longer yield zero recognized characters; 2. PDFs with missing font character mappings now use full-page OCR instead of garbled text; 3. Large knowledge base files are no longer killed midway through parsing; 4. Provider-default thinking is disabled for Qwen during parsing and image descriptions, and content moderation blocks now return a clear message; 5. Uploading a nonexistent file returns 400, knowledge base batch upload documentation no longer claims a false 50-file limit, and message attachments with missing IDs or invalid UUIDs return 400; 6. FAQ Excel imports skip example rows and deduplicate by question text; 7. Crawls that return zero pages without a retrievable error display “Unable to retrieve error details” and record additional metering fields; 8. The chat window is notified when attachment parsing produces no content or indexing fails, so attachments no longer remain stuck in processing; 9. When object storage uses a TWCA certificate, image and batch Q\&A file downloads apply the relaxed SSL domain settings so images can reach the model.
* **\[Build] Voice and scheduling fixes**: 1. Evaluation and scheduled jobs no longer exceed the queue's 30-minute acknowledgement timeout and restart the entire worker, also resolving stalled knowledge base processing; 2. When shared tools are skipped because their authorization has expired, schedules are marked “Partially completed” and their creators are notified; the Agent also explains that reauthorization is required.
* **\[Build] Text to SQL and database selection fixes**: 1. Date columns in Excel files uploaded to databases now use the date type, so queries “up to a given date” include the last day; 2. Table name casing, database names, PostgreSQL backticks, and colons in strings are corrected automatically before execution; column errors list the actual columns with a limit on how many are displayed; 3. Truncated table names in shared databases now match correctly, and reconciliation commands can restore orphaned tables and handle tables stuck for over 24 hours or failing deletion; 4. The AI assistant's database selector now lists all databases without pagination, fixing incomplete lists when there are more than 20.
Shared (21 items)
#### ✨ New features
* **\[Shared] Batch role permissions and member effective permissions**: Organization management now supports two permission governance capabilities: 1. Apply permission settings to multiple MaiAgent roles at once instead of editing each role individually; 2. View a member's effective permissions and see which role grants each permission. Permission screens have also been updated: all role permission names now use Chinese and match the interface, the role permission list is sorted by display name, Hook and conversation platform analytics permission translations are complete, and the quota self-check screen includes alert thresholds. [User guide](https://docs.maiagent.ai/org/roles)
* **\[Shared] Message Webhook signature verification and automatic retries**: Message event notifications (Webhooks) now include signatures and timestamps, allowing recipients to verify they come from MaiAgent and protect against replay attacks; the system automatically retries temporary failures. [User guide](https://docs.maiagent.ai/org/developer)
* **\[Shared] Role-based usage statistics visibility**: MaiAgent roles can now specify whether usage statistics cover the entire organization or only authorized AI assistants, with organization-wide visibility as the default. Organizations can adjust roles to let members see usage only for the assistants they manage. [User guide](https://docs.maiagent.ai/org/roles)
* **\[Shared] Organization reminder announcements and one-time prompts**: Administrators can send reminders to organization members with a notification that each member sees only once and that stops appearing after acknowledgement. This is suitable for maintenance notices or policy updates. [User guide](https://docs.maiagent.ai/org/overview)
* **\[Shared] Breeze-ASR speech recognition for on-premises deployments**: On-premises LocalAI now includes the Breeze-ASR-26 and Breeze-ASR-25 speech recognition models. LocalAI speech recognition settings are preconfigured at all three entry points (disabled by default), ready to use once enabled, with real-time pricing support. The default silence detection duration for real-time speech recognition has increased from 100 to 500 milliseconds to reduce interruptions during pauses. [User guide](https://docs.maiagent.ai/serve/voice)
* **\[Shared] Organization source IP allowlist**: Organizations can restrict login and access to approved source IP addresses. Once configured, features and APIs requiring authentication accept only allowlisted sources. The organization settings API also supports reading and updating the list, strengthening enterprise network access controls. [User guide](https://docs.maiagent.ai/org/overview)
* **\[Shared] Cohere Embed 5 Pro/Fast embedding models**: Knowledge bases can now use two additional embedding models, Cohere Embed 5 Pro and Fast, to find a suitable balance between accuracy and speed. [User guide](https://docs.maiagent.ai/build/knowledge-base)
* **\[Shared] More models in the model catalog**: Available models have been updated: 1. Claude Haiku 5.5 is available in the AWS Bedrock Japan region; 2. Qwen3.8 27B is enabled and will replace Qwen2.5 VL 72B, with assistants using the old model migrated automatically; 3. GPT-6.1 Sol via a direct OpenAI connection is listed ahead of the Azure version; 4. Muse Glimmer 30B is enabled; 5. Qwen3.6 27B is available (the same model shipped on-premises, pinned to DeepInfra fp8 without Chinese-funded providers); 6. DeepSeek V4.1 Flash on Azure is enabled, with a planned replacement by GPT-6 Luna in December 2026. [User guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Shared] Model retirement and replacement with automatic assistant migration**: The catalog is being updated: 1. A second batch of seven low-usage models (three Watsonx models, Mistral Medium 3.5, Qwen3.5 9B and 397B, and Nova 2 Lite) is retiring, with active assistants automatically migrated to replacement models; o3 remains scheduled to retire on October 20, 2026; 2. Direct OpenAI entries for GPT-5.6, GPT-5.5, and GPT-5.4 mini are retired in favor of Azure versions; 3. GPT-OSS 120B now uses Azure by default with the same model ID, and provider changes no longer retain the old API key; 4. Gemini 2.5 Flash is replaced by 3.5 Flash-Lite after retirement; 5. System background tasks select Gemini Flash according to the environment version (3.8 Flash in production); 6. Model menus now follow a fixed order: tier first, then newest to oldest within each tier (Claude: Fable, Opus, Sonnet, Haiku). [User guide](https://docs.maiagent.ai/build/ai-gateway)
* **\[Shared] Centralized model settings for system tasks**: The admin interface now groups system task models together. AI memory tasks use an explicitly configured model instead of automatically selecting the latest Gemini Flash, making costs and behavior more predictable. [User guide](https://docs.maiagent.ai/build/ai-gateway)
#### 🚀 Core performance improvements
* **\[Shared] Faster AI assistant response startup and completion**: Several parts of the response flow now spend less time waiting: 1. Embedding clients share a connection pool, eliminating one Bedrock connection handshake before the first model call; 2. Knowledge base results already ranked by the assistant's reranker no longer trigger another reranking call; 3. The AI assistant list no longer deduplicates across all fields, saving about 0.15 seconds per request; 4. Customer service conversation pagination counts no longer time out for non-owner members in large organizations; 5. The final message in non-streaming conversations no longer reverts to an earlier state; 6. API documentation schemas are cached for faster page loading.
#### 😊 User experience and interface improvements
* **\[Shared] More complete activity logs**: 1. Activity logs record additions and removals in an API key's enabled models list; 2. The exported API Path column is no longer blank; 3. Organization setting changes are fully recorded, and update events with no actual changes are omitted; 4. Conversation records for scheduled runs show credit consumption and processing time; 5. Conversation record exports include citation node IDs.
* **\[Shared] Minor improvements to chat platform and schedule settings**: 1. Partial chat platform updates no longer require the full theme configuration, and the zh-tw locale is stored as zh-hant; 2. Schedule timeout errors appear in the user's language; 3. Organization data includes the minimum schedule interval so the admin schedule form can indicate the lower limit; 4. Traditional Chinese translations consistently use 「身分」 instead of 「身份」 for identity.
#### 🔐 Security and Governance Enhancements
* **\[Shared] Stronger access isolation for conversations and sharing**: Conversation access is more tightly controlled: 1. Colleagues' MaiGPT conversations cannot be shared; sharing is limited to conversations the user can read; 2. Non-owners can only see their own MaiGPT conversations in lists, details, and counts; 3. Members cannot open colleagues' conversations from the personal chat interface; 4. Unauthorized callers cannot read customer service Copilot suggestion drafts, and drafts are also filtered out of existing shares; 5. Only conversation members can choose the Deep Research output format; 6. Updates to thinking settings return a response through the branch that actually authorized the caller.
* **\[Shared] Cross-organization access protection for attachments, schedules, conversation continuation, and Web Chat**: Closes cross-organization access gaps: 1. Conversation continuation and schedule creation only accept Agents and conversations within the caller's organization; 2. Schedules only accept Agents and chat platforms from the caller's organization; 3. Only attachments belonging to the current conversation can be linked, conversations in disabled channels of other organizations return 404, and non-object request bodies return 400; 4. Web Chat message sending and history are isolated by chat platform, with a data repair command provided; 5. Public registration is bound to the invited recipient.
* **\[Shared] Stronger login and authentication**: Login and invitation flows are more secure: 1. LDAP members can no longer bypass LDAP login through Forgot Password, and saving third-party login settings no longer turns Enable Login back on for owners; 2. The 30-day OAuth refresh token lifetime is enforced, with expired tokens cleaned up daily; 3. The OAuth authorization login page supports organization account (SSO) login; 4. Regular members can no longer create or copy organization invitations, and only owners can invite users to owner groups; 5. API key requests-per-minute limits also apply to endpoints with dedicated throttling; 6. API key authentication failures return 401 referring to Api-Key, and invalid keys return “Invalid API key”; 7. All four third-party authentication source types use the same naming rules (lowercase letters, digits, and hyphens, up to 31 characters), without affecting existing names; 8. When the registration page URL is not configured, invitations are marked as email delivery failures instead of silently remaining pending acceptance.
* **\[Shared] Logs no longer store complete keys; API request logs are retained for 30 days**: 1. Partner API and Webhook logs no longer store complete keys, and message Webhook headers are write-only; 2. Partner and general API request logs are retained for only 30 days and cleaned up automatically every night.
* **\[Shared] Stricter Credits balance checks**: Prevents credit balances from going negative: 1. Before a conversation reply, the wallet must cover the estimated credits for one reply; 2. Before parsing a knowledge base file, credits are estimated from the page count and parser, and parsing is blocked if the balance is insufficient; 3. API key and JWT callers undergo the same Credits checks as browser logins, and AI Gateway checks individual and role quotas; 4. Once insufficient credits are detected, failures in auto top-up or balance queries still return 402 and block the request.
#### 🛠️ Bug Fixes
* **\[Shared] Fixed AI reply interruptions, retries, and streaming errors**: 1. Bedrock stalled-stream retries replay clean messages that langchain can parse; 2. Redacted OpenAI reasoning content no longer interrupts Agent streams; 3. Agent model streams automatically retry if closed immediately after starting; 4. Azure/OpenAI rate-limit and temporary server errors reported during streaming are retried; 5. Replies interrupted by worker restarts or shutdowns appear as visible, unbilled failures; 6. Fixed trailing errors after stopping replies in on-premises deployments; 7. Hooks flush previously buffered text before tool items and at the end of the stream.
* **\[Shared] Fixed inaccurate Credits billing**: 1. Prompt cache write tokens for direct OpenAI connections are billed as LLM Cache Creation Token; 2. Replies with no output tokens still incur charges for recorded usage; 3. Unrecorded embeddings are no longer charged; 4. Replies completing video generation no longer charge the next turn's tool costs to the turn that sent the video; 5. Every Vision parsing run is charged, including reparsing, interrupted retries, and new file library versions; 6. When two settlements for an evaluation run concurrently, fixed fees continue from the previously billed question count, and judge calls are billed only once; 7. Platform-key usage by system features (automatic evaluation, AI image generation coordination, webpage extraction summaries, and document image descriptions) records cost only, with a charge of 0 credits.
* **\[Shared] Fixed assistant, organization, and AI Gateway settings issues**: 1. After a fallback group is deleted, its bound Agents use their own models for replies and billing; 2. Bulk member Metadata imports now merge data, preserving attributes and personal preferences not listed in the file; 3. The PII masking Hook template masks both cards when two cards appear consecutively, and licensed on-premises deployments include five system Hook templates; 4. On-premises deployments include missing database examples, and SynxDB is available when adding a database; 5. The Monday connector icon now points to a public CDN after the domain change; 6. In-app notification links use relative paths, so clicking the connector activation notification in on-premises and white-label deployments no longer takes users to the SaaS domain; 7. Non-streaming replies do not record output speed, while Agent mode replies record output tokens per second, preventing the model access page from displaying speeds of hundreds of millions of tokens per second.
## v2026.10.09
***
### 📌 Release Highlights
* **\[Serve] Fix conversation ownership across multiple Web Chats**: When multiple Web Chats are embedded on the same page and use the same user identifier (Source ID), creating a conversation with a contact from another conversation platform now uses the contact for the same user on the current platform instead, so the new conversation is created successfully and assigned correctly. This fix applies when both platforms belong to the same organization, a matching contact already exists on the current platform, and Source ID access verification is disabled.
### 🗂️ Full Updates (by Product Line)
Serve (1 item)
#### 🛠️ Bug Fixes
* **\[Serve] Fix conversation ownership across multiple Web Chats**: When multiple Web Chats are embedded on the same page and use the same user identifier (Source ID), creating a conversation with a contact from another conversation platform now uses the contact for the same user on the current platform instead, so the new conversation is created successfully and assigned correctly. This fix applies when both platforms belong to the same organization, a matching contact already exists on the current platform, and Source ID access verification is disabled.
## v2026.10.07
***
### 📌 Release Highlights
* **\[Build] Restored the Quick Reply creation option**: On the type selection page for adding an Agent UI, Quick Reply now appears alongside Flex Message and Quick Menu. You can open the existing editor directly to configure interactive reply buttons.
* **\[Build] Improved knowledge base retrieval after follow-up questions**: When you use an AI assistant in general mode and answer its follow-up question with a short phrase, such as a location, the system uses the conversation text to formulate the knowledge base search query. This reduces the chance of retrieving irrelevant content by overlooking the original question.
* **\[Shared] AI memory management restricted to your own memories**: “AI Assistant → Advanced Settings → AI Memory” now lists and manages only the signed-in user's own memories. The same restriction applies to organization owners. Viewing, editing, deleting, and bulk deleting memories are all protected, with no additional configuration required.
### 🗂️ Full Updates (by Product Line)
Build (2 items)
#### 🛠️ Bug Fixes
* **\[Build] Restored the Quick Reply creation option**: On the type selection page for adding an Agent UI, Quick Reply now appears alongside Flex Message and Quick Menu. You can open the existing editor directly to configure interactive reply buttons.
* **\[Build] Improved knowledge base retrieval after follow-up questions**: When you use an AI assistant in general mode and answer its follow-up question with a short phrase, such as a location, the system uses the conversation text to formulate the knowledge base search query. This reduces the chance of retrieving irrelevant content by overlooking the original question.
Shared (1 item)
#### 🔐 Security and Governance Enhancements
* **\[Shared] AI memory management restricted to your own memories**: “AI Assistant → Advanced Settings → AI Memory” now lists and manages only the signed-in user's own memories. The same restriction applies to organization owners. Viewing, editing, deleting, and bulk deleting memories are all protected, with no additional configuration required.
## v2026.10.06
***
### 🛠️ Bug Fixes
* **Fixed failed Bedrock response retries**: When a response stalls or times out while using an Amazon Bedrock model, the system now correctly resumes the response if an automatic retry has already returned an answer. This prevents errors in processing retry content from interrupting the response or affecting the next conversation turn.
* **Recovery after Canvas generation failures**: When creating charts or dashboards in Canvas mode, if the AI mistakenly uses a tool that cannot be executed, the system guides it back to the Canvas output flow. This reduces cases where only text is returned without generating a Canvas, as well as unnecessary point consumption from mistakenly using location search. This fix also covers the Canvas flow in assistant teams.
* **Fixed missing text at the end of Web Chat responses**: AI assistants with response masking or checking enabled now complete the checks and send the remaining text before finishing a response, preventing the ending from being omitted in Web Chat. Explanations preceding tool operations are also displayed in order, and existing content protection still applies.
* **Fixed the customer service entry disappearing after disabling transfers**: When an administrator edits a role that already has full access to customer service conversation pages, clearing “Transfer others' conversations” and saving now preserves the customer service conversation entry and existing page permissions. Selecting the parent customer service conversation permission also no longer automatically enables the transfer permission; it must be selected separately.
## v2026.10.05
***
### ✨ New Features
* **Minimum turns for multi-turn evaluations**: When creating a multi-turn conversation evaluation, you can specify the minimum number of turns required before it can end upon reaching its goal, preventing tests from stopping too early. The results page also shows this setting.
* **Wait for user replies in assistant teams**: Automatic and handoff connections in assistant teams can wait for a user reply before continuing, with optional reply choices. This supports workflows that need to confirm information step by step.
* **Default Credits allowance for new members**: Organizations can set the Credits allowance, expiration, and reset cycle for new members, and override the allowance when sending an invitation. When the settings dialog is reopened, previously saved results will not overwrite the new form.
* **Gemini real-time voice model**: Voice assistants now offer the Gemini 3.8 Live model for administrators to configure real-time voice interactions. Existing voice settings will not switch automatically.
* **New available model options**: The model menu now includes Claude Sonnet 5.5 (AWS Bedrock) and GPT-6.1 Sol (Azure). Administrators can select them for AI assistants and MaiGPT according to their organization's authorization.
* **Find files in the current conversation**: You can ask the assistant to list, filter, or reopen files in the current conversation, including uploaded images and assistant outputs, making it easier to retrieve previously used content across channels.
* **Cross-conversation AI memory for cloud organizations**: AI memory is now available to cloud organizations, allowing assistants to refer to saved preferences and facts in members' subsequent conversations. Memory is separated by organization and member; personal memory is not saved for visitors who are not logged in.
### 🚀 Core Performance Improvements
* **Improved loading of large knowledge base menus**: The knowledge base menu in conversations loads names and file counts first, fetching files and FAQ only when an individual knowledge base is expanded. Large collections of knowledge bases no longer load all content at once.
* **Improved preparation for knowledge base follow-up questions**: Knowledge base Q\&A skips unnecessary question rewriting. When rewriting is needed, it uses the platform's configured lightweight model and checks the result, reducing extra waiting before an answer begins.
* **Reduced waiting for concurrent usage deductions**: When multiple usage charges are settled simultaneously within an organization, price lookups, usage logging, and notifications no longer hold the wallet for extended periods, reducing contention between Credits deductions.
* **Improved conversation history loading**: When loading conversation history with citations and files, repeated retrieval of document content and download links is reduced, preserving the existing response content while reducing waiting.
* **Improved assistant memory initialization**: When an assistant starts answering, the auxiliary model used to organize memory is now prepared only when needed, reducing extra initialization before the first reply.
### 😊 User Experience and Interface Improvements
* **Role permission explanations and assistant list updates**: Role settings clarify the difference between AI assistant management permissions and conversation platform usage permissions to help administrators assign them correctly. The available assistant list is rechecked when members switch conversations.
* **Evaluation cost traceability**: Evaluations are billed for actual usage, including the models and knowledge retrieval used to answer, with costs aggregated per evaluation. Credits usage records can open the corresponding results directly; members without evaluation permissions see a permission notice.
* **Improved provider name display**: AI Gateway's credential menu for custom models and provider overview now show recognizable provider names, avoiding internal codes or blank names.
* **Customer service assignment and send reminders**: 1. Assignment and transfer lists show customer service agents' online status; 2. Unassigned conversations can be assigned directly to an agent; 3. Sending a message before taking over a conversation prompts a reminder, with options to take over and send, send directly, or cancel.
* **Meeting recording and upload checks**: Before live recording starts, microphone availability is checked, with prompts for permission or device issues. Offline uploads also check audio file content, blocking non-audio files whose extensions have merely been changed.
* **Automatic system announcement updates**: While the admin page is open, unread announcements refresh every 60 seconds. Notices automatically reflect the latest status when model services recover or announcements are cleared, without requiring a page refresh.
* **Meta channel setup guidance**: When creating a Messenger, Instagram, or WhatsApp conversation platform, the interface reads the current deployment's connection settings. Channels with incomplete setup remain visible and show notices about missing settings.
* **Duplicate assistant name notices**: When creating, copying, or renaming an AI assistant, duplicate names within the same organization are flagged below the name field, making it easy to correct the name before submitting again.
* **File deletion activity records**: Knowledge base file deletion events in activity logs retain the filename at the time of deletion, allowing administrators to identify the affected file after it has been removed. Mobile detail views now use a single-column layout.
* **Schedule quotas and completion status**: The schedule page shows the number of active schedules, the organization's limit, and the remaining quota. Schedules that reach their execution limit are marked completed, distinguishing them from schedules that still consume quota.
* **Conversation platform description field**: Conversation platform settings now include a description field where administrators can explain the assistant's purpose, giving organization-built assistants a clear introduction on marketplace cards.
* **Assistant team information masking explanations**: Node settings clarify attachment and paraphrasing rules when original user input is disabled, helping administrators understand what information each node can actually read.
* **Default reasoning effort behavior explained**: The default option in model reasoning settings explains what happens when no effort level is specified, and higher reasoning levels show a notice that waiting time may increase.
* **Login error notices within the form**: When an account or password is incorrect, the notice appears directly in the login form, with a forgot-password link to help users correct their input or reset their password.
* **Time awareness enabled by default for new conversation platforms**: New conversation platforms enable time-aware replies by default, allowing assistants to answer based on the current time and time zone. Existing platform settings retain their values, and the current time is also shown when editing.
* **Corrected file workspace operation descriptions**: The workspace selector now explains that assistants can write and perform file operations according to user permissions, removing wording that could misleadingly suggest the workspace is entirely read-only.
* **Provider deletion impact notices**: Before deleting an AI Gateway provider, the confirmation dialog shows how many models still use its credentials, helping administrators understand the impact first.
* **Consistent interactions in the Work workspace**: 1. Search includes a clear action, and dangerous actions such as deletion explain their consequences first; 2. Form submission displays errors together and focuses the first invalid field; 3. Conversations and groups support double-click renaming, with the latest result taking precedence during rapid successive renames; 4. Failure notices remain visible longer, and progress notices stay until the operation finishes.
* **File output cards in Web Chat**: When an assistant or a delegated sub-assistant generates files, Web Chat displays file cards so visitors can view and download the content from each output.
* **Adjusted greeting feedback actions**: Web Chat greetings retain the copy action and hide thumbs-up and thumbs-down buttons, which are excluded from satisfaction statistics, focusing feedback on actual answers.
* **Localized Hook template display**: When selecting a Hook template, its name, type, and description appear in the current interface language, helping administrators using Traditional Chinese understand its purpose.
* **Automatic reassignment when an agent pauses new assignments**: When a customer service agent pauses new assignments, automatically assigned conversations that they have not yet replied to are reassigned to other available agents, or returned to the queue if no one is available. Conversations already replied to or manually assigned keep their original assignment.
* **Improved Reply Formatting in Messaging Apps**: Assistant replies in LINE, Messenger, Instagram, and WhatsApp are converted to plain text so formatting symbols for bold text, headings, and other styles do not appear directly. Content with large numbers of repeated symbols also no longer stalls during format conversion.
* **Combining Consecutive Messages in Messaging Apps**: When visitors send multiple images in succession or send an image followed by text, the system waits briefly and combines them into a single response. Messages in the same conversation are processed in order to prevent simultaneous replies across multiple turns.
* **Trimming Older Tool Results for New Assistants**: Newly created assistants trim older tool results based on content size by default, reducing the space that historical tool output occupies in long conversations. Existing assistant settings retain their current values.
* **Automatic Replacement of Retired Qwen Models**: After Qwen3.6 Max is retired, the daily retirement job switches assistants using it to Qwen3.6 Plus and adjusts output length limits to avoid further calls to the disabled model.
* **Gemma Service Provider Change**: Gemma 4 31B is now provided through AWS Bedrock. The former OpenRouter version is registered for replacement, and the retirement job migrates assistants to the new provider so the old provider is not made available again after deployment.
* **Notifications for Expired Shared MCP Authorization**: When authorization for a shared MCP connector expires, the system skips that service and notifies organization members who can manage tools. In-app announcements and email guide them through reauthorization. Notifications are not repeated during the same authorization outage.
### 🔐 Security and Governance Enhancements
* **Sensitive Tool Information Masking Settings**: When editing a tool, you can specify sensitive headers to mask and view the platform's default masking entries, helping prevent authentication information from appearing in tool execution logs.
* **Model Origin Filters and Visibility Settings**: The model menu provides origin labels and filters, and organizations can centrally hide Chinese models. Model selection for web scraping also follows this restriction.
* **Permission to Transfer Conversations Served by Others**: Only members with permission to transfer other agents' conversations can transfer or take over a conversation that someone else is serving, preventing unauthorized assignments from interrupting ongoing support.
* **Stronger Meta Message Source Verification**: Incoming Messenger and Instagram messages undergo source verification. Deployments without the required verification information reject messages, preventing unverified requests from posing as channel messages.
* **Login Restrictions for Disabled Authentication Sources**: When an administrator disables login for an authentication source, Keycloak, EIP, SAML, LDAP, and linked third-party logins all honor the restriction, preventing login access through a disabled source.
* **Access Logs for Expired API Keys**: Unauthorized requests using retired API keys are recorded in the corresponding organization's API access logs, helping administrators trace connections still using old credentials.
* **AgentOps Monitoring Access Restrictions**: The AgentOps dashboard, conversation log export lists, and download information all require monitoring permission, preventing members with only other AgentOps permissions from reading organization monitoring data.
* **Preventing Forged Hook Verification Records**: Conversations no longer accept user-forged Hook execution results that could bypass personal data masking or input protection. Execution records are also no longer mixed into the assistant's fixed prompt content.
* **Stronger Identity Verification for LINE Web Pages**: When Web Chat is opened within LINE, visitor identity is accepted only after verification by LINE. LINE account information supplied by the web page itself is no longer trusted.
* **Stronger Activity Log Read Permissions**: Viewing activity log event details, export lists, and download links requires activity log permission, preventing unauthorized members from reading organization operation information.
* **Provider Response Storage Settings**: Responses requests sent directly to OpenAI or through Azure disable the provider-side response storage option, reducing additional storage of conversation replies through this interface.
* **Restrictions on Rejoining Removed Members**: After an administrator removes a member from an organization, SSO login and directory synchronization no longer add them back automatically. A new invitation is required to restore organization membership.
* **Stronger Information Masking in Assistant Teams**: When original user input is disabled for a node, attachment reading, parsing, and conversation log tools are also removed. Tool search no longer offers these restricted operations, preventing them from bypassing the masking scope.
### 🛠️ Bug Fixes
* **Knowledge Base Hybrid Retrieval Settings Fix**: When using a custom Elasticsearch knowledge base, connections supporting RRF are no longer incorrectly identified as unavailable, allowing hybrid retrieval to be enabled normally. A warning also appears if OpenSearch search configuration creation fails.
* **Deleting Connectors Together with Their Tools Fix**: When deleting a custom connector with the option to delete its tools selected, the deletion option is now sent correctly, preventing associated tools from remaining in the list.
* **Member Import Results Fix**: After batch imports of organization members or role members, results separately show counts for newly added and existing members, preventing existing members from being counted as successful additions.
* **Reply Streaming and Waiting Indicator Fix**: The first text segment of a reply appears as soon as it arrives, rather than briefly showing only a later segment. While a tool is processing and no text or cards are available yet, Web Chat continues to show a waiting indicator.
* **Duplicate Conversation Platform Sidebar Items Fix**: After opening a conversation from the support sidebar, the same conversation platform no longer appears multiple times due to pagination loading. Icons and channel labels also remain consistent.
* **Connector Authorization Status Fix**: Connectors using service accounts or requiring no authentication display an available status. They no longer ask members to authorize individually or show inapplicable deauthorization actions.
* **Admin Interface Acceptance Issues Fix**: 1. Meeting translation model searches now find matching names; 2. Usage statistics no longer accept future dates; 3. The delete button for the system administrator role is disabled in advance; 4. Activity logs no longer display inapplicable read events; 5. Access to the conversation export page follows monitoring permissions.
* **Tool Creation Dialog Fix**: When starting creation from a tool type card, the dialog no longer disappears immediately due to a page reload, allowing tool settings to be entered normally.
* **Gateway Trend Date Fix**: Trend charts in the AI Gateway overview display dates according to the selected time range. Queries spanning multiple days show both date and time so data from different dates does not appear to fall on the same day.
* **Knowledge Base Selection State Fix**: 1. Available knowledge bases update when search criteria change; 2. Rapid conversation switching retains the latest list, without older data returned later overwriting it; 3. MaiGPT can send messages directly when no knowledge base is selected and no filters are set.
* **Workbench Conversation Continuation Display Fix**: The Work workbench now displays context usage and continuation prompts for assistants other than MaiGPT and for admin test conversations. Creating a continuation conversation from a summary card keeps users at the original assistant entry point.
* **Disabled Skill Options Fix**: Skill menus for AI assistants, MaiGPT, and document quick actions list only enabled skills, preventing submission failures caused by selecting disabled organization skills.
* **Large Referenced File Preview Fix**: When opening large documents from knowledge base citations, file views, or monitoring, the preview wait limit increases from 20 seconds to 300 seconds, reducing timeouts before documents with hundreds of pages finish loading.
* **File Generation Failure Display Fix**: If an assistant is denied permission to start writing a file or writing fails, subsequent explanations remain visible as conversation text. They are not placed in a file card, and the file is not incorrectly shown as complete.
* **Connecting Existing WhatsApp Numbers Fix**: When connecting an existing WhatsApp Business App number to a conversation platform, the system correctly recognizes the coexistence connection completion event and no longer incorrectly shows registration as incomplete.
* **Role Section Preview Fix**: When only some sub-permissions are selected for a role, the section preview reflects the permissions actually available, matching the sections members can see after logging in.
* **File library version history entry fix**：The shared space root directory now includes an entry for version history. When switching workspaces or spaces, history stays separate for each scope, and older data that arrives later no longer replaces the current view.
* **Fallback display when maps are not configured**：When Google Maps connection details are not configured, the Web Chat map canvas displays an unavailable notice while keeping the location list and filters accessible, avoiding repeated loading failures.
* **Connector authorization status refresh fix**：Opening a conversation no longer triggers repeated connector status queries for the same identity. After authorization is completed, the latest status is fetched again so the connector no longer appears unauthorized.
* **Browser tool long operation fix**：When the assistant loads a slow webpage, it follows the configured wait limit, and long operations no longer lose the browser connection. Other conversations rejected because of usage limits no longer get stuck waiting for the browser to close.
* **Duplicate website assistant Widget fix**：Website assistants created through Ask AI reuse the assistant's built-in web Widget instead of creating a second identical entry. Settings customized by customers are preserved.
* **Delegated assistant interactive content fix**：When an assistant team delegates to a sub-assistant, the sub-assistant's question options, quick replies, rich cards, and files reach the user instead of being lost when results are handed back.
* **Video generation cost aggregation fix**：Conversation Credit consumption totals and conversation history exports now include video generation costs, giving administrators a complete view of consumption when checking usage.
* **Tool search authorization guidance fix**：When an assistant using tool search encounters an MCP tool that the contact has not authorized, it can still provide an authorization link, allowing the contact to complete the connection instead of only being told the tool is unavailable.
* **Keycloak connection error messages**：When SSO login fails because Keycloak certificate verification fails, a recognizable connection reason is displayed, helping distinguish verification configuration issues from general login failures.
* **Activity log recording fix for web operations**：Organization operations performed by administrators on the web are now correctly recorded in the activity log, fixing cases where an operation succeeded but its corresponding event was missing.
* **Existing conversation model synchronization fix**：When an administrator changes an assistant's model, existing conversations that inherit the assistant's settings automatically switch to the new model. Models explicitly selected by users within conversations are preserved.
* **Scheduled execution status fix**：Long-running scheduled tasks no longer start repeatedly. Tasks that exceed the execution limit are marked as failed, and running statuses left behind after unexpected interruptions are automatically cleared.
* **Provider credential preservation fix**：When editing a vLLM or OpenAI-compatible provider, changing only the name or other settings preserves existing credentials. Leaving the secret field blank no longer clears the API Key.
* **Automatic conversation evaluation connection fix**：When RAGAS automatic evaluation is enabled, consecutive scoring requests no longer log a connection error for each item because a shared connection has become invalid. Scores for each metric can be retrieved normally.
* **Tool failure reasons and trace fix**：When MCP or API tools cannot run because of insufficient permissions, connection failures, or other issues, the assistant and execution trace display understandable reasons. Failed calls are no longer marked as successful.
* **Model fallback failure handling fix**：AI Gateway fallback skips deleted self-hosted models. When no candidate model is available, it returns a clear failure reason instead of getting stuck on a model that no longer exists.
* **Blank response fix after team handoff**：After an assistant team hands off to the next node, the receiving assistant correctly identifies messages from earlier nodes. If its first attempt still produces a blank response, it tries once more.
* **Meta channel connection version update**：Messenger, Instagram, and WhatsApp connections and the Facebook login flow now use consistent, supported service versions, preventing channel failures caused by expired versions.
* **Presentation creation and image insertion fix**：Submitting through the presentation shortcut creates an editable presentation project. Generated images must actually be inserted into slides before completion is reported. Images not yet inserted are also available in the Insert Image drawer.
* **Model tool capability check fix**：Assistants using models that do not support tool calling no longer attempt to run tools. When deletion of a custom connector still in use is rejected, it is no longer incorrectly recorded as an authorization revocation.
* **Tool trace record link fix**：Tool nodes in execution traces now include the corresponding tool execution records, allowing administrators to review the inputs, outputs, and errors for each operation.
* **Assistant usage statistics fix**：Token usage from assistant responses is now correctly included in statistics. Hourly charts consistently group data by local time zone, preventing populated and empty time periods from being sorted incorrectly because of different time formats.
* **Knowledge base file move processing fix**：1. Files moved between knowledge bases are saved before processing starts, preventing moves from getting stuck; 2. evaluation report exports and test case generation follow the same order; 3. connector activation notifications are sent only after the application is saved, preventing missed notifications.
* **Disabled connector display fix**：Connectors disabled by the organization no longer appear among the connectors available in MaiGPT, including data returned by conversation updates and real-time messages.
* **Deleted fallback groups release credential references**：After a model fallback group is deleted, it no longer prevents the deletion of associated provider credentials. Unavailable settings are no longer counted as being in use.
* **Response emoji display fix**：When AI responses contain an index card emoji, it is displayed as ordinary text instead of being mistaken for a rich card and producing a loading failure message.
* **Bring-your-own-model usage recording fix**：When an organization disables Credits billing, assistant responses using models with its own credentials are still recorded in AI Gateway request logs, allowing administrators to review actual usage.
* **Traditional Chinese account message fix**：When using the Traditional Chinese interface, login failures and related account operation messages are displayed in Traditional Chinese instead of untranslated English prompts.
* **Duplicate monthly Credit allocation fix**：When plan changes and monthly Credit allocations run at the same time, the system rechecks the allocation time to prevent Credits from being added twice for the same month.
* **Claude conversation length calculation fix**：Available conversation length for Claude assistants is now calculated correctly according to model limits, preventing the first message in a new conversation from being incorrectly rejected as exceeding the length limit.
* **Video submission status confirmation**：The assistant must confirm the actual submission result from the video generation tool before reporting that a request has been submitted, preventing success claims before a generation job is created.
* **Bedrock response wait fix**：For conversations using Bedrock models with tool operations, waiting for the first response chunk and timeout handling are now consistent, preventing premature interruptions while the service is still preparing.
* **Azure non-GPT model response fix**：Non-GPT models such as Mistral on Azure now use compatible output length settings, fixing response failures caused by rejected settings after a conversation request was submitted.
* **Improved waiting after calendar synchronization failures**：After the first meeting calendar synchronization fails, the page reports the status and retries in the background. Reopening the page no longer gets stuck waiting for the same synchronization each time.
* **Office file output format fix**：When assistants generate Word, Excel, PowerPoint, or similar files, plain text is prevented from being written directly as a binary document, avoiding files that cannot be opened after downloading.
* **Conversation memory fix for human handoffs**：Visitor and human agent messages exchanged while customer service takes over are synchronized with the assistant. When control returns to AI, it can continue from the earlier conversation without overlooking information confirmed during the handoff.
* **Image Search Fix for Specified Filenames**: When users ask the assistant to find images by a specific filename or entity name, results are filtered by the original filename to avoid returning visually similar images unrelated to the specified subject.
* **LINE Authorization Link Preview Fix**: Previewing a contact's individual authorization link in LINE no longer consumes its one-time authorization eligibility prematurely. The authorization flow starts only after the user actually opens the link and confirms.
* **Parsing Fix for Special Presentation Charts**: When an uploaded PPTX contains unsupported 3D or other special charts, only those charts are replaced with placeholder content. The remaining slides can still be parsed, instead of the entire presentation failing.
* **Chinese PDF and Output List Fix**: When the assistant creates a Chinese PDF, it uses fonts that support Chinese characters and checks the finished file. The download list includes only the requested output files, without program files used during generation.
* **Traditional Chinese Preservation in Document Parsing**: In knowledge bases with Simplified-to-Traditional Chinese conversion enabled, text extracted from documents is converted to Taiwanese Traditional Chinese usage. This prevents Simplified Chinese content from appearing after text recognition in images embedded in PDFs. Code and links remain unchanged.
* **Thinking Setting Fix for GPT Reasoning Models**: When users connect directly to OpenAI GPT reasoning models and adjust reasoning effort, a compatible request method is used. This fixes the issue where specifying a reasoning level prevented the model from responding.
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.
Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:
```
GET https://docs.maiagent.ai/release-note/en/readme.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
