What’s new in Enconvo

Motion Films and Artifacts

• Added Motion films: ask for a short animated video and it plays in the panel and the Artifacts library, with a timeline of its scenes you can scrub, frame stepping, a scene list, loop and full screen.
• Added Export to Motion films: download a film as an MP4 video or a GIF from the panel, and the agent can take stills of any moment to check how its film looks.
• Added camera moves, a cursor that glides and clicks, hand-drawn marks, caption chips, word-by-word text, a spotlight, light effects and video clips to Motion films.
• Added a check the agent runs before it calls a Motion film done: it plays every scene out of sight and catches scenes that fail or draw a moment differently when played back.
• Added sound to Motion films: music the agent writes plays with the picture, each scene can be narrated in your text-to-speech voice with captions you can turn off, and MP4 exports carry the sound with captions saved beside the video or drawn into it.
• Added a rough cut to Motion films: the agent first lays a film out as labeled boxes for you to confirm before drawing it, picks from a few named looks, and can set its words in bundled typefaces that look the same in the panel, stills and exports.
• Added sharing and comments to Motion films: share a film the way you share a deck so people you invite see each new version, and leave a comment at a moment of the film that shows as a marker on the timeline and jumps there when pressed.
• Added thumbnails to the Artifacts library: it opens on a grid of pictures of each page, document, diagram, design, deck, dashboard and film, drawn on your Mac and kept per version, with a switch to the list that it remembers.
• Changed Motion films so asking the agent for a teaser, an explainer, a walkthrough or a social clip makes a film instead of generated footage, and its films open on a picture, keep their words inside the frame and large enough to read, and pace their scenes more evenly.
• Changed Motion films so the agent keeps a logo or a window steady across a cut, and puts a placeholder for you to fill where a date, price or name wasn't given instead of making one up.
• Changed Motion films so the agent calls a film finished only once each check warning is fixed or named to you with its reason, and claims nothing about what you offer that you didn't tell it.
• Changed Motion films so the agent shows you the rough cut and waits for your reply before finishing the film, unless you asked it to go straight on.
• Changed Motion films so a scene that takes over a quarter of a second to draw a frame stops and shows its name card, and the player keeps going.
• Changed Motion film exports to save into Downloads as they are made, so long films no longer hold the whole video in memory, and a film can now export up to 4 GB when the disk has room.
• Fixed the agent sometimes finishing a new Motion film of three scenes or more without showing you its rough cut first; it now waits for your reply on the rough cut unless your own words asked it to skip it.
• Fixed a Motion film's check warning that a frame shows nothing when it shows a small mark, such as a dot on a dark background; a warning the agent keeps in the film's last second now says it comes at the end.
• Fixed Motion film scenes that read the date with new Date() drawing differently each time; they now see the film's own time, as Date.now() does.
• Fixed a narrated Motion film's captions file going out of step with the picture after scenes were moved or removed.
• Fixed a Motion film export that would come out too large to save drawing every frame before it failed; it is now refused before it starts.
• Fixed Send to a friend on a Motion film sending an empty document; films can't be sent to a friend yet, so the button is gone for them.
• Fixed a design, deck or Motion film page that never finishes drawing leaving a background process running at full speed after a thumbnail, still or check gave up on it; the process now really ends, and a film's check names the scene that hangs.
• Fixed the Artifacts library's filter tabs scrolling sideways at the panel's width, which hid Films and the kinds after it; the tabs now wrap onto a second line.
• Fixed shared documents never getting a thumbnail in the Artifacts library; one whose sharing service can't be reached is drawn once it answers.
• Added a behind-its-source note to a design's Theme tab and a deck's Design system menu when the design system it uses has changed since, with a Sync button that brings it up to date.
• Changed a design to open with its Properties panel closed every time, so the canvas has the room, instead of remembering it open from the last time.
• Fixed design and film uploads taking a file whose contents aren't the picture, font or video its name says.
• Fixed design and deck uploads refusing a picture saved under another picture type's extension, such as a JPEG named .png, and SVG files with long comments before the drawing; the picture is kept as the type it really is.
• Fixed Chinese, Japanese and Korean text in decks exported to PowerPoint showing in a font without those characters; each line now names a font for its own script.
• Fixed a design, deck or dashboard an agent makes sometimes not opening beside the chat on its own while your Mac was busy.

Video Conference

• Added profile photos to conference calls: people who have one show it on their tile, in People, chat and captions, and while a friend call rings, instead of their initial.
• Fixed a conference call on a phone's browser pausing everyone's video for several seconds after coming back from the background.
• Fixed a conference call on a phone's browser sometimes not showing Reconnecting after coming back from the background.
• Fixed a voice-only conference call in a browser or the Android app staying silent after coming back from the background, when its connection still looked up but no longer brought anyone's voice; it now reconnects within a few seconds.
• Fixed a guest's end screen saying the host keeps the notes when the call was too short for any to be written.
• Fixed the call timer, notes times and Recordings entry of scheduled and repeating conferences counting from when the conference was set up instead of when its call began, so everyone on a call sees the same call time.
• Fixed the end of a call on the iPhone and Mac counting everyone on the call in its transcript: it now counts the people who spoke, and claims no transcript when nobody did.

Video Conference on the Mac

• Added a notes panel to conference calls on the Mac while Notes is on: the transcript so far, Ask AI, which summarizes the call and answers your questions about it, and your own notes and bookmarks (⌘B), which go into the call's entry in Recordings.
• Changed calls you start on your Mac to take notes and record audio from the start by default; change it in Settings › Video Conference, or turn it off in any call from Notes in the dock.
• Removed the Video Conference button from the top of the Chat Window; start a call from Video Conference in the sidebar.
• Changed the Dynamic Island's and the menu bar's Video Conference to open Video Conference's Home in the Chat Window instead of starting a call.
• Fixed the Mac's Conference window always showing dark: the page before you join and the pages after a call now follow light or dark appearance, while the call itself stays dark.
• Fixed a conference call's entry in Recordings showing different times from the call's notes panel when this Mac started recording after the call began; its player now plays each line from where it was said.
• Fixed the Mac naming your own iPhone or other devices among the people on a call, in the call's title and at its end: it now names only other people, each once, as the iPhone does.
• Fixed call notes that no computer wrote in time disappearing without a word: your Mac now says the call's notes couldn't be saved.
• Fixed a notes Doc being left behind for a call you had deleted from Recordings while its notes waited to be written.
• Fixed your own notes from a call disappearing after two days on a Mac that was off before they reached Recordings.
• Fixed call notes being written on another of your Macs on the same call instead of the Mac where you turned Notes on, which then saved its recording without notes.
• Fixed a click on a Video Conference card, such as an incoming call or Conference notes are ready, taking keyboard focus from the app you were using, so the next paste or keystrokes went to the card.

Video Conference on iPhone and Android

• Changed the call controls on the iPhone and in the Android app to one row of five round, named buttons on every call: Mute, Camera, Speaker, More and Leave, with Audio in place of Speaker on the iPhone while a headset is connected; during a video call the row rests over the pictures, and a tap on a picture brings it back.
• Added an Audio sheet to calls on the iPhone and in the Android app: hold Speaker to choose the phone, the speaker or a connected headset, with the system's own list last on the iPhone.
• Added a Camera sheet to calls on the iPhone and in the Android app: hold Camera to choose the front or back camera, mirror your video or hide your own picture, and tap Flip on your own picture to switch cameras.
• Changed More in calls on the iPhone and in the Android app to a grid of named buttons, such as Chat, Captions, Camera and backgrounds, and People; on the iPhone, More shows a dot when a new chat message arrives.
• Added Open notes on the iPhone's call: while notes are on, the Record and take notes card opens the call's transcript so far and Ask AI, which summarizes the call and answers questions about it with the iPhone's own model.
• Added TalkBack names to every button in calls in the Android app, including the Ready page's microphone and camera buttons and the pictures in chat.
• Fixed the times in an Ask AI answer on the iPhone's call sometimes not being links to that moment in the transcript.
• Fixed turning on notes in a call on the iPhone while its paired Mac is off or signed out: it now says the Mac isn't reachable instead of turning notes on with nothing to write them.
• Fixed the Notes Doc sometimes not opening from the iPhone call's wrap-up when notes were turned on from the iPhone; the Mac the iPhone is paired with now writes them.
• Fixed Flip camera on the iPhone sometimes leaving the camera on the same side while your camera was on.
• Fixed a camera on the iPhone that couldn't start staying on with no picture: it now turns off and says the camera couldn't start.
• Fixed a guest's name in a call's People list on the iPhone sometimes showing only "…".
• Fixed the Android app leaving the phone's call sound on the speaker after you left a call, so a call in another app could then start on the speaker.
• Fixed the Android app's call window always showing dark: the page before you join and the pages after a call now follow the app's Appearance, light or dark, along with the status and navigation bars, while the call itself stays dark.

iPhone and Android

• Added Run in the cloud for iPhone chats on Enconvo Cloud models: the reply keeps going in Enconvo Cloud after you leave the app or lock the phone, and is waiting when you come back, with iPhone tools such as your location answered on your return.
• Added Motion films to the iPhone and Android apps: a film's card shows in the chat that made it, the Artifacts library has a Films filter, and a film plays on the phone with scrubbing, frame stepping, mute, captions and full screen, and exports to the share sheet.
• Added calling a friend from the Android app: the video button in a friend's chat, Call on their page, a friend's or teammate's name in Video Conference's search box, and Call again under Earlier ring them, as on the iPhone; during a call, Call rings them into it, and the call's Add people lists your friends and Team.
• Added incoming calls on Android: when a friend or teammate calls you, your Android phone rings with a call notification, also with the app in the background, and Answer joins the call straight away; a call you didn't answer leaves a Missed call notification.
• Added Sign in with Apple to the iPhone app: sign in with your Apple Account, or link and unlink it in Settings › Account › Sign-in methods.
• Changed Mac chats on iPhone and Android to show their messages with the Mac's own chat view, built into the app, so background tasks, Subagents, and steering look and work as they do on the Mac, and Find in Chat marks matches in it.
• Fixed Try again in a forked chat on iPhone and Android, which also deleted that reply from the chat it was forked from.
• Fixed the Android app closing on Android 12 and older when you typed a search into the side panel's address bar, and search tabs and history there showing the site's name instead of your search words.
• Fixed the light around the screen in iPhone Live Talk so it rises and falls with each word again, both while you talk and while Enconvo answers.

Friends and Friend Chat

• Added sending to friends from agents: ask Mavis or any agent, on the Mac or from your phone, to message a friend or teammate by name, or send them files or an Artifact through Friend Chat, and it asks you to approve each send.
• Added managing friends from agents: ask Mavis or any agent to read a Friend Chat, save a friend's file, add, remove, block or unblock someone, answer a friend request, mute, pin or clear a chat, or call a friend, and it asks you to approve each one.
• Changed the approval an agent asks for before it messages, calls or changes a friend to show who it is for and what it sends, so you can check it before you allow it.

Agent Phone

• Added keypad presses to calls an agent places to other people through Twilio or Telnyx: when an automated phone menu asks to press 1 or enter an extension or a number the call's brief gives, the agent presses the keys itself.
• Fixed keypad presses on calls an agent places through Twilio or Telnyx not reaching automated phone menus, and the agent entering the same keys twice before the menu had answered.
• Fixed calls an agent places to other people hanging up on an automated phone menu that says "press any key to continue", taking it for voicemail; the agent also now presses keys only once the menu has offered them, and no longer switches language when a menu offers another one.

Projects, Ideas and the Board

• Added a team workflow to every Project with members: Standard plans each goal, builds it, then reviews or tests it, and the Project's Workflows tab lets you pick who plans, who builds and who reviews, turn the check on or off, and set how many fix rounds a task gets.
• Added giving a Project's planning to a member: the member writes the plan as soon as you ask, and the lead then proposes its tasks as a Work plan that waits for your Start.
• Added choosing a team workflow, such as Ship or Planner, to run every goal of a Project while its custom workflows are on; the lead no longer picks Ship or Planner for a single task unless you name it with @.
• Added New team workflow to a Project's Workflows tab, which makes a team workflow with plan, build and review steps for the Project and opens it.
• Added a Custom workflows switch to Project settings, and a Turn workflows on button in the note a workflow shows when it is opened in a Project whose workflows are off.
• Added Ask the lead to set it up to Project settings when Find it finds no check command: the lead's chat opens with the request ready to send.
• Added the Sandbox switch to a Project's Lead chat: its Permissions popover now turns on the Sandbox for the whole Project, so the lead and each Enconvo Agent's tasks there run their commands sandboxed, while coding agents keep their own.
• Changed Full access in a Project's Lead chat and its tasks to work as it does in any other chat: it no longer asks before deleting, sending, publishing, paying or pushing, and a plan's card no longer lists approvals up front at Full access.
• Changed a Project lead with no members to do the work itself in the chat, with no Work plan or Board tasks, until you add members.
• Changed a Project whose lead has no members to say so in its Chat, Board and message box: the lead does the work itself, so the Board has no tasks and there are no members to name with @.
• Changed the empty Board of a Project whose lead has no members to picture the lead doing the work in the chat.
• Changed a Project's Board to show the tasks of the goals you give the lead in Chat or Ideas: it no longer has New task, its cards group by idea with each idea's title opening its chat, and the tasks of a goal you gave in Chat group under that chat, after the ideas.
• Changed an idea you start to work toward its goal: the lead's plan takes the idea's goal and Done when line, and the idea moves to Done only once the lead judges that goal met.
• Changed an idea's Start to Plan it in a Project with members, since the lead plans the tasks first and you start the plan.
• Changed Ideas in a Project whose lead has no members: Start hands the idea to the lead, which does the work in the idea's chat and moves it to Done when it's finished.
• Added kinds to Ideas (Feature, Bug, Improvement, Docs, Research): a new idea gets one guessed from its words, you can change it from its card, the New idea dialog or the idea's page, and the Ideas board filters by kind.
• Changed a kind Enconvo guessed for an idea to show as a dashed chip on its card in Ideas, which says Kind guessed from your words on hover.
• Added Where it stands to an idea's page in Ideas, which shows Plan the work, Do the work and Review & test with how far each has come and who does it.
• Changed an idea with no goal of its own to show its plan's goal and Done when under Outcome, marked From its plan.
• Fixed the @ menu in a Project's Lead chat offering workflows while the Project's workflows are off.
• Fixed a plan proposed before a Project's workflows were turned off still following its workflow after Start; its card now says it runs without it.
• Fixed a task left out of a plan following its Workflow after Workflows were turned off in the Project.
• Fixed the Project lead trying a task again on its own when the task's worktree couldn't be made or its Prepare command failed, which only ends the same way until the setup is fixed.
• Fixed sending more to a failed or stopped Board task skipping what Try again does; it now counts as a new try, the tasks waiting for it wait for it again, and its finished plan runs again.
• Fixed a Work plan started while the team is paused showing as started though nothing ran; it now starts paused with the rest and runs when you resume.
• Fixed the Project lead offering to answer a task's question or add to a finished task for you, which it can't do; it now points you to the task's card.
• Fixed the Project lead saying only that a plan waits for Start when Enconvo couldn't start it on its own; it now says what holds it.
• Fixed stopping a Board task just as its run ended, which could end it twice and show it as failed or done after the Stop.
• Fixed a Project's Board and Ideas staying on their loading screen when Enconvo couldn't read them; they now say so and offer Try again.
• Fixed opening a Project's Board right after the Chat Window opened showing the old Orchestrator Board instead of the Project's own Board.
• Fixed Additions the Project lead sent to a Board task being labeled Orchestrator; they now show the lead's name.
• Fixed lines that name a Project's lead, in the Workflows tab, Ideas, plans and the questions waiting for you, leaving a blank when the lead Agent has no name.
• Fixed the Project lead planning more tasks after a Planner task, which already hands its tasks back to you as a new plan.
• Fixed a workflow's approval step offering Allow once and Deny on its Board card, in the chat's Needs you list, and in the iPhone and Android Project panels; it now offers Approve and Decline.
• Fixed Reset on a Project's Standard team workflow also turning off the team workflow chosen to run every goal; it now only puts Standard's steps back.
• Fixed a Project's Workflows tab still showing a member who had left as the one who plans, builds or reviews; that step now falls back to its default.
• Fixed a workflow chosen to run every goal that later became a task workflow, or was deleted, silently doing nothing; Standard now runs each goal and the Workflows tab says why.
• Fixed a Project's lead following Standard's planner, builder and reviewer picks on top of the team workflow chosen to run every goal.
• Fixed a Project's lead being told to give the planning, building or reviews to a member whose Agent is paused or whose coding agent needs an update, which its plan then refused; that step now goes to the default until the Agent is back.
• Fixed a Project's planning member starting to plan on its own when you hadn't asked, such as after a plan ended; it now starts at once only when you ask, and otherwise waits for your Start.
• Fixed a Project lead with no members skipping the report on a plan that had just ended.
• Fixed a review task in a Project looking for the work it checks in the Project's folder instead of on that work's own branch.
• Fixed a Project member's task that writes a file, such as a document, working in your own checkout and switching its branch; it now gets a worktree of its own like a coding task.
• Fixed Ship and Planner asking you to approve an idea you already started from Ideas.
• Fixed the Folder access question a plan with Codex asks as it starts ringing your phone when nobody answered it, and staying open after the plan ended.
• Fixed the folder picker for Use an existing folder in New project so it has a New Folder button.
• Fixed an idea finished a moment ago reading "done now ago" on its card in Ideas; it now says done just now.
• Fixed Ideas' All and Needs you counts ignoring the kind you filter by, and Add idea hiding while a kind is picked; an idea you add then starts as that kind.
• Fixed the New idea dialog keeping the kind of an Ideas filter you had since changed; a kind you pick yourself stays.
• Fixed a Project's Board no longer grouping by idea after you hid a lane or changed what its cards show, and an idea group's tooltip saying Open idea where it opens the chat.
• Fixed a Board you never picked a grouping for keeping the grouping it had when you hid a lane, instead of following its default.
• Fixed an idea that finished when you merged its work showing the time its task ended as when it was done.
• Fixed a long idea's short title sometimes coming back in another language than the idea's, such as German for an idea written in English or Chinese.
• Fixed an idea in In progress showing no status after its tasks were done; it now says when its goal isn't met yet or the lead is still checking it.
• Fixed an idea you reopened staying in In progress after you merged a task that had been waiting for review.
• Fixed a done idea keeping the Needs you highlight and count, and Mark done on an idea already done moving it in the Done list.
• Fixed Edit on an idea putting back its long first line when the short title arrived while the dialog was open.
• Fixed an idea moving to In progress while a member was only planning it; it now stays in Planning until the plan's tasks run.
• Fixed an idea the lead finished itself, with no Work plan, staying In progress in a Project with members; the lead can now move it to Done.

Workflows

• Added a Steps view to the workflow editor that lists a workflow's steps from top to bottom with their conditions, repeats and parallel steps, and keeps the canvas one click away.
• Added an Add menu between steps in the Steps view that adds an agent, a prompt, an approval, a workflow, a condition or a repeat right where you want it, and a job's Add a step that also adds a step that runs at the same time.
• Added a line in the Steps view to each step that runs only after an earlier step with a condition, such as Only if “Notify” runs, since a skipped step skips the steps after it.
• Added a Team workflow or Task workflow choice to New workflow, which you can change later from the strip under the workflow editor's header.
• Added two team templates to New workflow, Plan, build, review and One member, offered when Team workflow is chosen; a team workflow made from them or described to the assistant has the Member slots a Project needs to use it.
• Added Add members… to a workflow's Member slot picker, since only a Project's members can fill a slot.
• Changed the Steps view to pick steps with ⌘-click or Shift-click as the canvas does, and to show the bar for picked steps at the bottom, where it no longer covers the first line.
• Changed the Steps view to read a job's condition and the jobs it waits for with each of its steps in VoiceOver.
• Changed a Repeat until line in the Steps view to say the step fails if its condition still doesn't hold after the last time.
• Changed the Steps view to take one Tab stop, the selected step with its menu and Add, with the arrow keys moving between steps, and fixed the place to drop a dragged step staying on a step the pointer had left.
• Changed the Steps view to say an Otherwise never runs while the step it stands in for is turned off, and stopped suggesting Always run for an Otherwise, which is skipped with that step either way.
• Changed workflows so a step that runs once per item stops starting new items after one fails, unless the step is marked as not required; in Ship, the tasks after a failed one no longer run.
• Changed a workflow step that repeats until a condition holds to say what its last round found when it runs out of rounds, such as the check's last lines or the failures verification listed.
• Changed Sprint and Ship so the Address step runs only when the review found something that needs fixing, and Sprint checks the project again only when code changed after its first check; findings left for later still appear in the closing note.
• Changed the Planner workflow to need a Project in git, like Ship and Sprint.
• Changed Compound to apply the lessons you kept to the repo's docs and rules.
• Changed Compound's approval to say that Approve applies every lesson, and that you can keep only some by declining with the ones to keep and trying the task again.
• Changed the Mac notification for a workflow's approval step or a coding agent's approval to offer Approve and Decline instead of Approve and Deny.
• Fixed Try again after you declined a workflow's approval forgetting what you wrote; the new run's steps are now told what you declined and why.
• Fixed workflow steps being told to commit their changes while their instructions said to leave them for the next step's review.
• Fixed Verify setup looking finished whether or not its runbook passed; its steps now show whether the proof passed, and a proven check command becomes the Project's check when the Project has none.
• Fixed workflow steps holding a task with only an architecture design to screen rules, such as matching copy strings and making every screen reachable.
• Fixed the workflow editor marking a Project Ready, and letting Run start, a workflow that builds on a branch in a Project whose folder isn't in git; it now says the Project needs a git folder.
• Fixed the Workflows tab not saying when reading its list again failed; it now says so with Try again.
• Fixed a Project's Workflows tab showing Loading every time you came back to it, and its whole list looking dimmed while workflows were off.
• Fixed Ship and Planner showing as task workflows in the workflow editor, and losing their team scope when exported or pasted as YAML.
• Fixed adding, copying or deleting a step in the workflow editor making a later step without an id read another step's output.
• Fixed the Steps view showing jobs that wait for different jobs as At the same time, and Starts right away on a job's first step that still waits for its job.
• Fixed the Steps view's Only if line naming the wrong step for an Otherwise and for a step after a job with a condition, and showing one on a step set to always run; after an approval it now says Only if you approve “…”.
• Fixed the workflow assistant's chat showing its whole instructions instead of your description when it builds a new workflow.

Dictation and Draft

• Changed the Dictation capsule into a smaller pill with the live transcript in a soft bubble above it; rest the pointer on the pill while you dictate and it opens into a card where you can switch the microphone, turn what you said into a Draft, or pick a Clean up option, without leaving the app you're typing in.
• Added Clean up progress to Dictation: after you stop, the pill and the bubble show the option at work on your words, offer Insert as Spoken when it takes longer than usual, then show that it's done, or that it failed and your original words went in instead.
• Changed the Draft panel to the Dynamic Island's look: it sits at the bottom center of the screen just above the Dictation capsule instead of next to the text field, grows smoothly as the draft is written, scrolls a long draft inside the panel, shows the Draft keys as separate keys, and says whether ↵ replaces, inserts or appends once that is decided.
• Changed the Draft panel to show the draft as the model writes it instead of waiting for the whole text; ↵ pressed while it is written puts it in once it is done, and a slow model now has 30 seconds to start writing rather than to finish.
• Added refining a draft with the Dictation key alone: whenever the Draft panel shows, also while a draft is being written or after no instruction was heard, the Dictation key records a new instruction for it instead of typing into the app under the panel, and a Dictation hotkey that is a single key, such as F5, no longer closes the panel.
• Added continuing with Draft: ask it to continue or to add something at the end, such as a closing line or another paragraph, and it writes only the new text and puts it after your selection or the field's text instead of rewriting it.
• Added Prompt for drafts to Settings › Dictation › Draft: read and edit the instructions Draft follows to write and change text, and Restore default puts back the prompt that comes with Enconvo.
• Added a one-time Draft introduction on the Dynamic Island after your first dictation: it shows the keys to hold, and Show Me opens the Guide tour's Draft step to try it.
• Changed the Hold ⇧ to draft hint to show only once, the first time you dictate over selected text, and not at all once you have used Draft.
• Fixed Draft changing people's names when asked for a more formal tone, such as writing a Chinese name in Latin letters or adding a title to it; names now stay as written unless you ask to change them.
• Fixed the previous draft lingering as a faded copy in the Draft panel when you press the keys again to change it: the panel now clears at once and listens for your change the way a new draft does.
• Fixed Draft saving each instruction and draft as a hidden chat, which Desktop Sync copied to your other computers when it was on; drafts now stay out of your chat history.
• Fixed Dictation post actions, such as Polish, Translate or an AI Command, saving each dictation and its result as a chat that Desktop Sync copied to your other computers when it was on; they now stay out of your chat history.
• Fixed Dictation and Draft sometimes reading from or pasting into a window left on screen above the app you were typing in, such as a stale Force Quit window or another app's floating panel, instead of the app in front.
• Fixed Dictation with no microphone connected: instead of a paused notice, the HUD now says it is waiting for a microphone and keeps Reconnect ready for when one is plugged in.
• Fixed a leftover transcript bubble from the last dictation flashing above the Dictation HUD when the next dictation starts.

Recording and Captions

• Fixed Recording's live captions losing the last words said right before a pause or Stop, along with their translation.
• Fixed Recording's live captions showing one person's speaker number on someone else's lines after the connection drops.
• Fixed a recording whose live captions dropped saving the last caption before a pause or Stop with only part of its translation on a slow connection.
• Fixed a recording whose live captions dropped keeping only part of a caption's translation when speakers took quick turns or one spoke at length.
• Fixed a recording with two-way English and Chinese live translation that dropped saving reworded Chinese speech as its translation.

Clipboard History, Bookmark Search and File Search

• Changed Clipboard History, Bookmark Search and File Search to native Mac lists that open faster and follow the system's look, in the Smart Bar and in a window of their own; their hotkeys open them as quickly the first time after launch as later, the loading light runs under the field for every search, and pressing Return in Clipboard History now pastes almost instantly.
• Added a Hotkey Opens a Window setting to Clipboard History, Bookmark Search and File Search: on by default, their hotkey opens the list in its own window and leaves the Smart Bar as it was, and turned off it opens the list in the Smart Bar; the switch is on each plugin's page in Settings, and each list's Actions (⌘K) and the Smart Bar's ⋯ menu open that page.
• Added an Alias to Clipboard History, Bookmark Search and File Search, and they now show in Smart Bar search by default: type the alias and a space (then what to look for) to open the list in the Smart Bar.
• Fixed the Shortcut on the Settings page of Clipboard History, Bookmark Search and File Search so it opens the native list (in a window of its own by default) instead of an older web version, and the Alias there now works the same way.
• Fixed @-mentioning Clipboard History, Bookmark Search or File Search in a chat so the agent searches that list.

Chats and Agents

• Added the Sandbox switch to the Permissions picker in the Smart Bar, the Dynamic Island and the App Sidebar, under the approval modes as in the chat window, and the Permissions button shows a small box while the Agent's Sandbox is on.
• Changed the Chat Window, Settings and chats on the iPhone and Android apps to open faster: pages that open on a click, such as Plugins and Settings, now load in the background once the window is up, and code colors load after the page shows.
• Changed Codex chats to stop showing Codex's warnings, such as under-development features or config notices, in the conversation; they are still listed under the /warnings command.
• Fixed a new chat's title sometimes coming out in another language than your message, such as Chinese for a message in English.
• Fixed the /, @ and other menus opening while you type a URL or path in the chat box, the Smart Bar, the Dynamic Island and the iPhone and Android apps: they now open only at the start of a message or after a space.
• Fixed the Chat Window coming back on a new chat after its page reloaded, such as after the page crashed; it now reopens the chat it showed.
• Fixed plugin pages in the Chat Window sidebar, such as Email, going blank when opened again after another plugin page: each page now comes back as you left it.
• Fixed archiving a chat with Desktop Sync on: a chat archived on one Mac no longer comes back to the chat list on your other Mac, or on the Mac that archived it, when the chat finishes a run or changes there before the archive arrives.

AI Commands

• Changed Explain to answer in the language of the selected text by default, with a language of its own to pick on its AI Commands page.
• Changed Explain to treat every send on its own, without earlier messages, so each answer follows only the text you just sent.
• Changed selected and pasted text to reach the model as plain text, once, so AI Commands use fewer tokens on each run.
• Changed Response Language on the AI Commands page to sit in its own Language card instead of under Advanced.

Computer Use

• Fixed Computer Use screenshots of a window with an alert sheet open showing the sheet stretched over the whole window, so places read off the picture missed the sheet's buttons; the picture now shows the window itself, and the sheet's buttons are clicked by their index.
• Fixed Computer Use describing Finder's view buttons and other toggles in icon, column and gallery views as number values, which also made the agent's note after a view switch say those buttons had moved.
• Fixed Computer Use saying a panel that hides while its app is in the background, such as Finder's View Options, may have closed, which led the agent to press the panel's shortcut and close it; it now says the panel is hidden and comes back when the app is active again.
• Fixed Computer Use clicking a control its window doesn't show, such as Finder's icon size slider while the status bar is off, and reporting success although the click reached nothing; it now says the control is outside the app's windows.
• Fixed Computer Use sometimes clicking another element than the one its index named, such as a Finder sidebar item instead of the icon size slider, on the first click in an app after Enconvo started.
• Fixed Computer Use typing long text into a web page, such as Enconvo's own chat box, sometimes putting parts of it out of order, and saying only part of the text landed when it typed into an empty web editor.
• Fixed Computer Use's Command shortcuts, such as Select All, Paste and Undo, doing nothing in web pages such as Safari's and Enconvo's own chat box while saying they worked, which also kept its paste option from inserting any text.
• Fixed Computer Use's Select All doing nothing in a text field of an app in the background, such as TextEdit, so the text it typed next was added to what was there instead of replacing it.

Image and Video Models

• Added OpenAI's GPT Image 2.6 Flare and Sunburst to Image Create, with your own OpenAI API key, your ChatGPT sign-in, or Enconvo Cloud; OpenAI image providers without a saved model now start on 2.6 Flare, and the iPhone and Android apps list both models.
• Added the GPT Image 2.6 controls to OpenAI image generation: a complexity setting, markup-guided edits that follow a transparent PNG you draw on, and SVG and animated GIF output with a frame count and frame duration.
• Added GPT Image 2.6 Flare and Sunburst through PipeLLM in Image Create, with the complexity setting, markup-guided edits, and SVG and animated GIF output.
• Added an output format choice (PNG, JPEG or WebP) to OpenRouter image models that offer one, such as GPT Image 2.6 Flare and Sunburst on OpenRouter.
• Added Gemini Omni 1.2 Flash to Video Create as the default Google Gemini video model, with videos from 3 to 15 seconds and a choice of 360p, 720p, 1080p or 4K.
• Fixed GPT Image through PipeLLM refusing a 3:1 or 1:3 aspect ratio, and ignoring a 2K or 4K resolution asked for without an aspect ratio.

Calendar and Reminders

• Fixed Apple Calendar all-day events created or changed by an agent ending a day early.
• Fixed agents changing or deleting one occurrence of a repeating Apple Calendar event by its day, or one that was already changed.
• Fixed agents being unable to stop an Apple Reminders reminder from repeating, and made a repeating reminder without a due date, or a repeat sent in the wrong form, come back with a clear explanation.

Desktop, Smart Bar and Island

• Changed the stop button that shows while a reply is running so it no longer looks like a screen recording control: in the Smart Bar it is now a round button with a square, drawn like the Send button, and in the Dynamic Island it reads Stop with its Esc key.
• Fixed the Live Desktop board staying still after it opened: it now updates as your agents start, work and finish.
• Fixed chat previews on the Live Desktop board and in the Dynamic Island showing the model's reasoning instead of its answer.
• Fixed screenshots copied to the clipboard sometimes going missing on paste: a screenshot tool's region drag no longer triggers the PopBar's selected-text copy, which put the previous clipboard back over the new image.
• Fixed Enconvo freezing for up to a few seconds when it read the app context of a document window while that app was slow to answer, such as Xcode while it indexes a project.
• Fixed the app context of a long file open in VS Code and editors like it holding only the part around the cursor while a screen reader or another accessibility tool was on; it now holds the whole file.

Settings

• Added renaming paired phones in Settings › Connections: a pencil beside the phone's name, so an Android phone listed as "Android" or an iPhone listed as "iPhone" can be told apart.
• Fixed opening a plugin's settings from another window while Settings is already open, which landed on the Plugins page instead of that plugin's page.
• Fixed French, Korean, Portuguese and Traditional Chinese text in some places switching to an informal or regional wording, such as Brazilian Portuguese or mainland Chinese terms, instead of the formal style and local terms the rest of the app uses.

Projects and Work Plans

• Changed @ in a Project's Lead chat, on the Mac and the iPhone: naming members no longer sends your message straight to them; Mavis, the lead, plans the work and gives each member you named its part in a Work plan, and files can now go with the message.
• Changed a Work plan for members you name with @ in a Project's Lead chat or a chat with Agents to start right away, so you no longer press Start after every message that names who should do the work; a plan whose member can't run yet still waits for Start.
• Added @ that brings an Agent into a Project: naming one of your Agents with @ in a Lead chat adds it to the members, with a line you can undo, and gives it its part, unless you turn off Anyone you @ joins in the header's Members view; in a Project without members, the first Agents you name become its members.
• Changed a Project's members so one is enough and there is no limit of six: add as many Agents and Coding agents as the Project needs, in Project settings or from the Lead chat's header.
• Added a Lead picker to the header of a Project's Lead chat, so you can change the Project's Lead and add, remove or give roles to its members right from the chat, as an ordinary chat lets you pick its Agent.
• Changed Projects made without members so they work like any other: their chats have the Lead and Members picker with Chat and Board, a Board, and the start page with the Lead's card, and the last member can leave from the header.
• Added the Project's Board to a Lead chat's start page: before the first message, cards for Needs you, To review, Running and New results take the members' place, clicking a card lists its tasks under it with Allow once, Decline, Try again or Review (clicking it again folds the list), and a task opens the Board right at it.
• Changed a Project's lead so it does the work no member fits itself, the Project's files included, without stopping after a set number of steps and handing the rest to a member.
• Changed Work plans in a Project whose folder isn't a git repository: the tasks that change files now take turns in the folder on their own, while tasks that only read run beside them, and the plan's card shows which task each one waits for, before and after Start.
• Changed the Work plan card to a simpler layout: each task is one line with its member on the right and its Done when line beneath it, Approve up front folds into one line, and the Permission mode and cost share the footer beside Start.
• Added task chips to Lead chats and task chats: each Board task number, such as T-65, shows the task's state and name (the number alone when the reply names it again), hovering previews it, a click opens its detail in the side panel beside the chat, and ⌘-click opens it on the Board.
• Added automatic retries for Board and Work plan tasks that a passing AI model error cuts short, such as an overloaded or rate-limited model, a server error or a dropped connection: the task runs again in its chat after a pause, picking up where it stopped, up to three times, its card says while it waits, and it fails and notifies you only when the model keeps failing.
• Removed the "You allowed … once" lines that piled up above a Lead chat's message box after you answered its members' approvals.
• Fixed a long command in a Lead chat's approval card filling the window: it now scrolls in a box of its own height.
• Fixed a Claude Code or Codex task in a Work plan handing on its chat's warnings, such as Codex's MCP startup notice, and its notes from before its last step together with its answer, so the task's card, the Lead and the tasks after it get only the answer; Paste and Copy on a Coding agent's reply leave such warnings out too.
• Fixed a forked chat's copy of a Work plan card starting or changing the plan of the chat it came from: the copy now says to start and change the plan there, with Open chat.
• Added a way to point a Project at its folder again after the folder was moved or renamed: Project settings marks the missing folder, and Change… picks its new place, keeping the Project's chats and worktrees.
• Fixed pointing a Project at its moved folder when a worktree made inside it (such as Claude Code's claude -w) moved along: git now works in that worktree again, and the worktree Enconvo took in from it follows to its new place instead of showing as missing.
• Fixed a Claude Code or Codex chat in a Project, including one that was replying at the time, refusing new messages with "The folder no longer exists" after the Project was pointed at its folder's new place; the chat now follows the folder and goes on with its conversation.
• Fixed Claude Code forgetting what it remembered about a Project after the Project was pointed at its folder's new place.

Ideas

• Added Ideas to Projects with members, beside Chat and Board: write down what you want the Project to do and press Return, or talk it through with the lead first, then start an idea to have the lead plan it in a chat of its own and follow its tasks on the idea's card, from Not started through Planning and In progress to Done.
• Added dragging ideas in Not started into your own order, also with Option-Up and Option-Down, as with To start's tasks on the Board.
• Added saving an added idea again from its chat: when you and the lead change what it asks for there, the lead updates that idea on the Ideas board instead of writing down another.
• Added an idea draft's card to its chat that stays in view after the lead's turn ends and when you open the chat later, so a draft you left when the chat beside Ideas closed can still be added, added and started, or dismissed, and it no longer offers Add once the draft is dismissed or merged.
• Changed Board cards in a Project to name the idea their task came from instead of labeling every planned task as the lead's idea.

Chats Outside a Project

• Added Work plans to chats outside a Project: Mavis can hand parts of a bigger request to your other Agents and Coding agents in a Work plan, never by starting them straight from her or her Subagents, and every task works in the chat's own folder, where the tasks that change files take turns.
• Changed chats outside a Project so Mavis gives work to your other Agents only when you name them with @; otherwise she does the request herself.
• Added your other Agents to the @ menu of chats outside a Project, on the Mac and in the iPhone app: naming one there has Mavis give it its part of the request in a Work plan.
• Added a Board to chats outside a Project: the Chat | Board tabs in their header open one Board with the tasks your Agents work on across all those chats, each card naming its chat, and Start, Stop or a follow-up on a card acts in that chat.
• Added a start page to new chats outside a Project: under "What can I help with?" it shows the tasks that need you, wait for review, are running or just finished, or, when there are none, every agent you can chat with in one row that scrolls sideways, the ones you chat with most first, where clicking one names it with @ in your message, as a Project's member cards do; a coding agent whose CLI isn't installed stays out of the row.
• Changed the start page of a new chat so its Board, member and Agent cards ease in one after another once they load, and picking a Board card unfolds and folds its tasks smoothly.
• Added Start without and Give this task to on the Work plan card of chats outside a Project, on the Mac and in the iPhone app: a task whose Agent is paused, deleted or not signed in says why, with Sign in where it helps, Start without starts the rest, and before you press Start a task can go to any of your other Agents, those already working in the chat listed first.
• Fixed a chat outside a Project refusing a Work plan or Give this task to for a new Agent once six of your Agents had worked in it; Agents with no task waiting or running there now make room.
• Fixed tasks given out by a chat outside a Project saying they came from "New Chat": they now name the chat by its title, also after you rename it.
• Fixed deleting a chat that had moved from outside a Project into one leaving the Work plan it proposed there waiting or running on the Project's Board.
• Fixed a chat moved into a Project after giving tasks to your other Agents still offering them in its @ menu and working outside the Project's Boundary.
• Fixed Mavis adding a review to a one-Agent Work plan in a chat outside a Project once other Agents had worked in that chat.

The Board

• Changed the Board's look: an empty Board now shows a picture of how tasks reach it with one way to start (New chat on the Chats Board, Add task elsewhere), each lane ends at its last card instead of running down the whole Board, on a fainter background, with an empty lane keeping room for a card and saying what it waits for, cards are filled instead of outlined, with the task's number and age at the top, its Agent at the top right and a clear state mark, and keep their other actions under a More button in the corner, so a narrow card shows its details without spilling over, and a task's details open in a Liquid Glass panel whose steps show glossy marks, with Open chat and Stop up top, a failed task's Try again, Lend another Agent and Skip under its title, and the rest under More.
• Changed the Board to open at once when you come back to it, showing it as you left it while it catches up, and to start loading as the pointer rests on a Project's or the Chats Board's Board tab; a first open shows its lanes instead of "Opening the Board…".
• Added arrow keys and Linear's single keys to the Board: the arrows move between cards, up and down in a lane and across to the next lane with tasks, the card they reach shows a blue edge and halo, an open task's details follow the card, and closing them puts the keys back on that card; Space opens or closes the card under the pointer, C starts a new task, Option-Shift-Up and Option-Shift-Down move a To start card to the top or the bottom, and dragging the Board's empty space scrolls its lanes sideways when they don't fit.
• Added lane options to the Board, as in Linear: hovering a lane's header shows its menu, where Hide lane moves the lane to a Hidden lanes column at the right that brings it back with a click, and turning off Show empty lanes moves lanes with no tasks there too; in a narrow window that column folds to a strip that lists them, a hidden Needs you with tasks waiting for you glows amber there, To start's header also gets a + for a new task, and each Board remembers how you set out its lanes.
• Added Display to the Board, as in Linear: it turns Show empty lanes on or off and picks what cards show (Agent, number and age, status, buttons, workflow, Planned by, reviews, waiting Additions, chat, branch, commits and attempt), its button counts what is off, Reset to Default turns it all back on, and each Board remembers it.
• Added picking several Board cards at once, as in Linear: X picks the card under the pointer or the one the arrow keys reached, Shift-X or Shift-click picks every card from the last one picked, Command-click adds or removes one, and a floating bar starts, merges, stops or drops the picked tasks each action applies to, names in its tooltip the ones it leaves, asks once more before dropping, and Esc clears the picks.
• Added Group by Agent to the Board: each lane gathers its cards under their Agent, with the Agent's face, name and count, and each Board remembers the choice.
• Added a note when the Board's filters or search leave every lane empty, with Clear filters to bring every task back.
• Removed the Board's tasks from under their Project in the Chat Window sidebar, so it lists only your own chats and worktrees: follow the tasks and open their chats on the Board.
• Fixed the Board's member faces: the chosen member now shows a blue edge, and a narrow window hides the faces that don't fit instead of cutting one off.
• Fixed the Board's search box and a task's edit fields losing their edge while you type in them; they now show a blue edge.
• Fixed the Board in a narrow chat window cutting off New task and the filters: the base branch, New task's label, the running slots and then Project settings step aside, and the filters drop their counts before one is cut.
• Fixed Tab on the Board's filter row shifting it sideways or stopping on member faces hidden in a narrow window; the Board's buttons, filters, faces and New task's choices, and the buttons on task rows in chats, now show where Tab is.
• Fixed opening a Board task on the Board from elsewhere opening the Board without the task: it now opens with that task's details showing, where its Merge and Drop are, also when the Board is already open on another task, and a filter or search that hides its card clears.

Board Tasks

• Fixed a Board task whose worktree folder was deleted outside Enconvo: trying it again or adding to it now brings the folder back from the task's branch with its work, instead of its agent failing or working in the Project's own folder.
• Fixed the Board saying a task's branch was gone when only its worktree folder had been deleted; it now shows the branch's commits, and never reads changes from a git repository above that folder, such as a home folder kept in git.
• Fixed asking a Board task for more in its own chat making Mavis work in the Project's own folder instead of the task's worktree.
• Fixed asking a Board task for more in its own chat after its worktree was removed, because the task changed nothing or the folder was deleted, working in the Project's own folder or failing: the task now gets its worktree back first.
• Fixed a Claude Code or Codex Board task's chat refusing a new message with "The folder no longer exists" after the Board removed the task's worktree because it changed nothing; the worktree now comes back first, in the same folder even when the task was renamed since.
• Fixed trying a Board task again, or adding to it, after the Board removed its worktree because it changed nothing: the round now gets a new worktree first instead of running without one.
• Fixed a Board task that a restart cut, after its worktree folder was deleted, being told another branch's commits, or none, instead of its own when it ran again.
• Fixed asking a finished Board task more in its own chat making its member work in the Project's own folder instead of the task's worktree, and that worktree being removed while the member worked in it.
• Fixed a Board task whose request named a file by its path in the Project's folder being told both to change it and never to touch that checkout; the task now changes its own worktree's copy.
• Fixed a Board task merged outside the Board, such as with git in a terminal, keeping its worktree and branch on disk with no way to remove them from the sidebar: the Board now removes them once it sees the merge, as its own Merge does, and clears ones left behind this way before; a worktree with uncommitted changes stays.
• Fixed a Board task merged with git outside Enconvo keeping its worktree on disk until its Board was opened: Enconvo now notices the merge within a few minutes and removes the worktree and branch, as the Board's own Merge does, including the worktrees earlier versions left behind after such a merge.
• Fixed Board tasks merged by hand, or sent another round after a merge, moving to Done while their worktree still held work nobody had merged; they now stay in To review until that work is merged too.
• Fixed new work on a Board task merged by hand staying hidden under Done after a follow-up in its chat: the task goes back to To review.
• Fixed a Board task whose pull request was merged on GitHub losing the worktree changes it hadn't committed, and waiting for an open terminal on its chat before it finishes.
• Fixed a Board task whose pull request was merged on GitHub while the Project's check or a member's answer still ran in its worktree losing that worktree under them, which failed the merged task when the check sent it back.
• Fixed a Board task whose pull request was merged on GitHub while the Project's check ran starting a fix round on the merged pull request when the check failed.
• Fixed a Board task that failed or hit a usage limit staying in Needs you after its work was merged; it now moves to Done, and adding to it starts a new task.
• Fixed a merged Board task offering Remove worktree again and again when only files a dev server wrote back were left in its folder.
• Fixed the Board's task drawer saying a task's leftover worktree holds uncommitted changes when it was kept for another reason, such as a pull request squashed on GitHub; it now says what removing it keeps.
• Fixed the Board removing the worktree of a task merged by hand while the Project's check, or a new round of the task, was still running in it.
• Fixed the Board removing a finished task's worktree while its Coding agent chat went on in a terminal, or while a Project member's question was being answered in the task's chat.
• Fixed the Board removing a finished task's worktree folder when its branch had taken commits since the Board last looked, which left the task pointing at a folder that was gone.
• Fixed a Board without a Base branch treating a task as merged when you only checked its branch out on a branch of your own, which removed the task's worktree, its ignored files included, and its branch.
• Fixed the Board counting a worktree that an agent made inside a task's worktree as the task's uncommitted changes, and removing that inner worktree even when the task's own changes kept its worktree.
• Fixed the Board removing a task's worktree, with files written meanwhile, when the task started another round or files were written there while the removal was still checking the worktree.
• Fixed a Board worktree removal that was called off because its task started another round still removing the worktrees made inside it, and leaving behind a saved copy of the task's uncommitted changes each time it was tried.
• Fixed Merge, Open pull request, Remove worktree and Drop on the Board committing or removing a task's work while its Coding agent chat went on in a terminal: they now ask you to quit the agent there first.
• Fixed the Board starting another round in a task's Coding agent chat while that chat went on in a terminal: sending a task back and Try again now ask you to quit the agent there first, and a pull request's failing checks wait until it ends.
• Fixed a Board task's new rounds (a lead's send-back, a review's findings, a failed check, a lent Agent or a workflow step) starting beside a Coding agent that was continued in Terminal from the task or one of its step chats; they now wait or say why until the agent there ends.
• Fixed Merge, Drop and Remove worktree going ahead on a Board task that started another round while they were checking for an open terminal, which could commit the round's half-written files or remove its worktree.
• Fixed Merge on the Board taking in commits a Board task made after Merge was pressed, such as a new round's half-done work, when the merge had to wait for other work in the Project's repository.
• Fixed Stop on a Board task while it was still getting ready (bringing back its worktree or running the Project's prep command) letting the round's agent work anyway, and leaving the prep command running in the background.
• Fixed stopping a Board task right after an Addition or a retry still running the Project's prep command for up to 10 minutes, and a workflow's check step going on after its task was stopped.
• Fixed an Addition to a Board task, or a failed check's fix round, starting a round in a task that changed while the Addition was checking its chat, such as one whose pull request was merged meanwhile, and a Lead's send-back or a fix round starting in a Work plan you had just paused or stopped.
• Fixed pressing Try again twice in quick succession on a failed or stopped task starting two rounds at once in its chat.
• Fixed a Project member's long-running answer in an ended task's chat losing that chat after 20 minutes, which let a second question join the run and let the Board remove the task's worktree while the answer still worked there.
• Fixed Try again on a Board task that follows a workflow ending at once with an old run's result when its previous turn failed before it started, and a workflow step refused by an open terminal leaving an empty step chat behind that later skipped the task's brief.
• Fixed a Board task that follows a workflow leaving its workflow run going after its round failed, such as when it was to pick the run up again after Enconvo reloaded and a terminal was open on one of its chats, so that Try again ran a second copy beside it.
• Fixed a workflow step of a Board task that was stopped just before its member started leaving out the task's brief and your added words when the step ran again, and a Board task's worktree being merged or removed while a member answered in its chat after Enconvo reloaded the plugin.
• Fixed a Board task that follows a workflow losing, when Enconvo reloaded the plugin mid-run, the words you had added for its later steps, and running two of its steps at once.
• Fixed the chat details card of a Board task's chat, and of its workflow steps' chats, showing the Project's branch and changes instead of the task's worktree: it now names the worktree, its branch against the Board's Base, what it changed, its pull request and its ports, Open in opens the worktree, and it updates while open when the task merges, opens a pull request or changes state, or when you come back from editing in another app.
• Fixed a Board task whose worktree was deleted by hand still showing that worktree in its chat's details card when a dev server wrote its cache back into the folder.
• Fixed a Board task whose branch took in its Base, such as after resolving a merge conflict in its worktree, counting the Base's commits and files as its own on the Board and giving those files to the Project check as the task's.
• Fixed Board cards and task rows showing a code block's language, such as "text", in place of the first line of a task's result.

Worktrees

• Changed where worktrees go: a Project's chat worktrees and its Board tasks' worktrees are now together in ~/.enconvo/worktrees/<Project>/, each named after its worktree or task; ones made before stay where they are.
• Added Archive to a Found worktree's row in the sidebar, so a worktree another tool made can leave the list without being deleted: it moves to Archived worktrees, its folder and branch stay as they are, and Undo shows it as Found again.
• Changed the chat details card to name the Main checkout when a chat works in a Project folder that has worktrees, and to show a worktree made outside Enconvo, such as Claude Code's, under Worktree instead of Workspace.
• Fixed the sidebar's Main checkout row keeping a Project's old branch name after the branch was switched by a chat or in a terminal: it now follows the branch the message box shows.
• Fixed a chat whose worktree was archived, deleted or is on another computer listing under the Project's main checkout in the sidebar: it now sits in a Worktree unavailable group after the worktrees.
• Fixed a main checkout chat's details showing Changes +0 −0 when the uncommitted work was new files: their lines now count, as they do for a worktree.
• Fixed Archive merged worktrees skipping a worktree whose pull request was merged on GitHub while a chat was working in it: it is now archived when the chat finishes, unless the chat added work after the merge.
• Fixed Start from offering to continue the Base branch itself, which made a worktree whose Finish merged into itself and whose changes never showed: the Base branch is now only offered as the branch a new one starts from.
• Fixed Start from listing only the 30 newest open pull requests: typing now finds older ones by title, branch or number.
• Fixed switching the main checkout's branch offering a branch a worktree holds when the Project has more than 200 branches, where git then refused the switch: every branch a worktree holds is now marked In a worktree.
• Fixed a worktree's branch sometimes keeping its placeholder name when the agent was committing at the moment Enconvo named it.
• Fixed an archived worktree that couldn't be restored because Finder had left a hidden file in its old folder; when something else really is there, the message now says what to move and offers Reveal in Finder.
• Fixed a worktree whose folder was deleted while a dev server kept running in it: it now shows as missing instead of letting chats run in a folder without the Project, and Archive and Delete work for it again.
• Fixed restoring a worktree whose folder was deleted when the worktrees folder is reached through a symlink, such as a ~/.config kept in a dotfiles folder.
• Fixed a worktree another tool made and locked showing in the sidebar as a Found worktree after its folder was deleted or its drive was put away, and restoring one you had archived while its drive was away, both of which gave its chats a folder that wasn't there.
• Fixed a worktree made outside Enconvo that isn't on a branch offering New chat and Archive in the sidebar that always failed: its row now says No branch and how to use it.
• Fixed Claude Code, Codex and other Coding agent chats in a worktree running dev servers and commands without the worktree's own ports, so they no longer clash with the main checkout's or another worktree's.
• Fixed Claude Code, Codex, Grok Build and Antigravity chats in a worktree changing the Project's own copy of a file when asked about a path in the Project's folder; they now change the worktree's copy and leave the Project's checkout alone.
• Fixed a fork or Side chat of a Claude Code, Codex, Grok Build or Antigravity chat in a worktree listing under the Project's main checkout, without the worktree's ports, while it worked in the worktree.
• Fixed a Claude Code session you ran in a worktree's folder opening from the new chat page as a chat of the Project's main checkout; it now opens in its worktree, with the worktree's ports.
• Fixed a Claude Code session you ran in a worktree made outside Enconvo, such as with claude -w, opening from the new chat page as a chat of the Project's main checkout, or of the worktree whose folder it sits in; the worktree is now added to the Project and the chat opens in it.
• Fixed Continue in Terminal on the new chat page's session list resuming a Claude Code session you ran in a worktree in the Project's main checkout, where it changed your own copy of the files; it now resumes in its worktree, with the worktree's ports.
• Fixed Continue in Terminal and Copy Command for a Claude Code, Codex or Grok Build chat in a worktree leaving out the worktree's own ports, so the terminal's agent ran its commands on the Project's default ports and could change the Project's own copy of a file.
• Fixed archiving or deleting a worktree while one of its Claude Code, Codex or Grok Build chats was open in a terminal: the agent there kept working in the removed folder and could break Restore; the confirmation now says so, and archiving quits that agent first, while a merge, a pull request merged on GitHub, or archiving the worktree's last chat from the sidebar keeps such a worktree instead of archiving it.
• Fixed a new worktree copying the worktrees Claude Code keeps inside your Project (in .claude/worktrees) when Files to copy names their folder, which made a large copy and let git commands in it change the other worktree's branch; they are now left out.
• Fixed archiving or deleting a worktree that has another worktree inside it, such as one claude -w made there, deleting that worktree's uncommitted changes and leaving git listing a folder that's gone; a clean one is now removed properly with its branch kept, and one with uncommitted changes, or one open in Enconvo with chats of its own, keeps the worktree, and the Archive dialog and Delete name it before you confirm.
• Fixed committing or merging a worktree that has another worktree inside it, such as one claude -w made there, committing that worktree as a broken submodule entry; it's now left out, and it no longer counts as an uncommitted change.
• Fixed the Works in menu counting a worktree claude -w made inside your Project as an edited file of the main checkout.

Agents

• Added seven built-in agents beside Mavis in Agents and the agent picker: Coder, Designer, Reviewer, Researcher, Writer, Planner and Advisor, each with its own role instructions, with the Designer working on the design canvas with your design system or Enconvo Crystal.
• Added a Built-in badge on the Agents page for the agents that come with Enconvo, so they are told apart from the ones you made.
• Changed the agents that come with Enconvo so they can be paused but no longer deleted, since deleting one was permanent across updates with no way to get it back.
• Added an Edited or Default chip on each instruction field of a built-in agent (Instruction, User Message, CHARACTER.md and IDENTITY.md) with a Reset to default button that asks before it discards your text.
• Added a notice on a built-in agent's instruction field when an Enconvo update changes the default text you edited, with Use new default and Keep mine.
• Changed clearing an instruction field of a built-in agent to put the default back when you leave the field, instead of saving empty text that the default silently replaced, and the field says so while it is empty.
• Changed the instruction fields on an agent's settings page to save a moment after you stop typing, and when you switch tabs or leave the page, instead of on every keystroke.
• Changed the Reset configuration confirmation of an agent to say what goes back to its defaults (instructions including CHARACTER.md and IDENTITY.md, model, voice, memory and behavior) and that its chats, files and paused state stay.

Answers in Chat

• Added interactive apps to chat answers on the Mac: when an answer needs real logic, such as a simulator whose sliders recompute a chart and totals, a filter that regroups figures, a game, an animation or something to drag and draw, the Agent now writes a small app with the chat's own components and theme, light and dark, styled the way you ask when you ask for a look of your own.
• Added a Fix button to an interactive app that fails in chat, which shows what went wrong and asks the Agent to repair it.
• Changed interactive simulations you ask for to explore an idea to play right in the chat answer instead of opening as an artifact; an app you ask to build and keep still opens as an artifact.
• Added diagrams to chat answers: when an answer explains steps, a decision with branches, the layers of a system, a cycle or a phased plan, the agent draws it as a clean diagram right in the conversation.
• Changed answers about messages passed between parties, states and their transitions, or how data relates to draw a real sequence, state or relationship diagram in the chat instead of squeezing it into a step-by-step flow.
• Added tree diagrams to chat answers: an org chart, a taxonomy or another hierarchy now draws each part under the one it belongs to, joined by lines and colored by branch, and shows as an indented list in a narrow window.
• Added practice quizzes to chat answers: pick an option, check it in place, and see why it is right or wrong.
• Added flashcards to chat answers: cards to memorize, such as vocabulary or terms, turn over with a tap one at a time.
• Added checklists to chat answers: a packing list or the things to do before an event shows as items you tick off in place, with a count of what is done, and the ticks stay when you open the chat again.
• Added ticking off task lists in chat replies: a list of boxes the agent writes in its reply can be ticked in place too, and the ticks stay when you open the chat again.
• Added step-by-step walkthroughs to chat answers: when you ask to go through something one step at a time, step through it with Previous and Next, or jump to a step from the progress bar.
• Added tabs to chat answers: a workout program or other plan by day shows one day at a time under a row of tabs, each day's exercises a list you tick off as you go, instead of a list or table for each day.
• Added schedules to chat answers: a day planned hour by hour or a week of study shows as a calendar with each block as tall as it lasts, colored by kind with each kind's hours added up, and as a list by day in a narrow window.
• Added option cards to chat answers: when you ask which of a few plans or products to pick, they show side by side with their prices, what each has and lacks, and the best fit marked, also when you ask in Chinese or another language.
• Changed answers that compare a few options you're choosing between, such as plans, to show them side by side in a table with the best fit marked instead of a list.
• Added maps to chat answers: when an answer is about places, such as the stops of a day trip or a trip planned day by day, they show numbered on an Apple Maps map with a route line and the list under it.
• Added funnel charts to chat answers: stages such as a signup or sales funnel draw as narrowing bars with how many each step keeps and loses and the steepest drop marked, and the answer's text names the same steps and figures.
• Added scatter charts to chat answers and the iPhone and Android apps: two measures of the same items, such as study time and exam scores, draw as points on two number axes, colored by group, with a fitted trend line and its correlation when asked, where the Agent drew a picture by hand before.
• Added bubble charts to chat answers and the iPhone and Android apps: when each item has a third measure, such as a product line's revenue beside its growth and margin, the points become bubbles sized by it with each one's name on or beside it and the range of sizes in the legend, where the Agent drew a picture by hand before.
• Added heatmaps to chat answers and the iPhone and Android apps: one measure across two categories, such as calls by weekday and time of day, draws as a grid of cells shaded by value, with the busiest cell named and a low-to-high scale, where the Agent built a table colored by hand before.
• Added Gantt charts to chat answers and the iPhone and Android apps: tasks that each run from a start to an end, in weeks, days or dates, line up as bars on one timeline so overlapping work shows at a glance, colored by team or phase, where the Agent drew a diagram with made-up calendar dates before.
• Added waterfall charts to chat answers and the iPhone and Android apps: how a starting figure becomes a final one, such as revenue to net profit, draws as a bar per step from the running total, increases green and decreases red, with subtotals and the total standing from zero, where the Agent built the bars by hand before.
• Added Sankey charts to chat answers and the iPhone and Android apps: amounts that flow from where they come from to where they go, such as income into spending and savings, draw as bands as wide as what they carry, each destination in its own color, where the Agent drew boxes with the amounts written in before.
• Added calendar charts to chat answers and the iPhone and Android apps: the days something happened, such as a month of workouts or a habit, fill in on a month grid so streaks and gaps show at a glance, and daily amounts shade deeper on busier days with the highest day named, where the Agent built a table of dates before.
• Added treemaps to chat answers and the iPhone and Android apps: how one whole splits into parts, such as disk space or a budget by department and line item, now shows as tiles as large as each part, a department's items in its own color with each tile's amount and share written in it and each department's total under the chart, where the Agent drew horizontal bars or a crowded flow chart before.
• Added target lines and marked points to line charts in chat answers and the iPhone and Android apps: a goal, an average or a limit draws as a dashed line with its value, and a launch, a peak or a function's vertex and roots get a ring on the line with a short note, where the Agent drew a flat extra line or wrote the note into a month's name.
• Changed charts in chat answers to a cleaner look, with several series on one chart, a legend, values on hover, and line, area, pie and donut charts.
• Changed calculators and choices in chat answers: each slider shows its name and value above a wider track, a choice among a few options shows as buttons that wrap with the picked one filled, and a choice can show the picked option's details.
• Changed the Agent to ask what it needs before a task, such as your preferences for a trip or a purchase, as questions in the chat that you answer by picking options or typing, several in one go, instead of a list of questions in its reply.
• Added picking several options in a chat card, such as the names you like, and sending your picks to the Agent, on the Mac and the iPhone.
• Changed the Agent to offer candidates you pick from, such as names, as options you tick and send back, instead of a list in its reply.
• Fixed Agents set to Auto answering without tools when a request needs a card, such as options to pick from, questions to answer first, a comparison or a quiz, so the card didn't show.
• Fixed options you pick in a chat card showing as blank pills in the Chat Window, and changed picked options to a light outlined fill, so the card's send button stands out, on the Mac and the iPhone.
• Fixed visual answers that sometimes showed "This view couldn't be shown" instead of the answer: small markup slips, such as a tag left open, are now mended as the view draws, earlier chats included, and the agent draws again any view still broken.
• Fixed visual answers that sometimes came out as boxes of loose text or an empty table instead of the timeline, table or map they were meant to be, and maps that showed a failed step before they appeared.
• Fixed chat answers that sometimes showed a card's raw markup as text instead of the card or a plain answer.
• Fixed chat answers that sometimes added a box of plain text, such as a diagram's title or its code, beside the answer itself.
• Fixed chat answers repeating under a visual answer what it already shows, or only pointing to it: the text now ends with what the view doesn't say, such as the takeaway or what to do next.
• Fixed sequence diagrams in chat answers that showed only their code when a message or note contained a semicolon.
• Fixed state and sequence diagrams inside a chat answer's card showing as code: they now draw as diagrams.
• Fixed architecture and layer diagrams in chat answers that were sometimes written as Mermaid code instead of drawn as the chat's own diagram.
• Fixed the start and end capsules of step diagrams in chat answers showing only their name and dropping details such as a count.
• Fixed practice questions on a topic with math that sometimes came as text to answer in a reply instead of a quiz you check in place, and worked calculations that were sometimes drawn as a step diagram instead of formatted math.
• Fixed plans laid out by day or by time, such as a trip itinerary or a study week, that sometimes came back as a table for each day instead of a timeline or a calendar.
• Fixed progress bars in chat answers stopping at 100% when an amount is past its budget or target: they now show the real percentage and mark the part beyond the target, and the iPhone and Android apps show the real percentage too.
• Fixed planned-versus-actual figures, such as spending against a budget, that sometimes came back as a text table instead of bars showing which items are over.
• Fixed charts in chat answers that left the unit, such as the percent sign, off their axis and values when the Agent named it only in the axis or series name.
• Fixed chart values and sliders in chat answers showing a space before K, M, B or T, such as "$42 K".
• Fixed amounts in chat answer calculators showing extra zeros, such as 2.00 eggs in a recipe scaled to your servings.
• Fixed tables in chat answers running past the card's edge when a cell held a long note: cells now wrap to fit.
• Fixed tables in chat answers that showed empty, or made the Agent redo its answer, when it gave the rows and columns as data; they now show on the Mac and on the phones.
• Fixed Chinese and Japanese tables in chat answers wrapping one character to a line in a narrow chat: each column keeps room for a few characters, and a table too wide for the chat scrolls sideways.
• Fixed headings of groups in chat answers, such as each day's title in a trip plan, that went missing on the Mac and on the phones.
• Fixed figure tiles in chat answers that cut off a long label; the label now wraps, and the figures beside it stay in line.
• Fixed colored words in chat answers showing as raw HTML tags on the Mac and in the iPhone and Android apps: they now show in a color that reads in light and dark themes.

Chats and the Chat Window

• Added a tray above the message box listing what a chat has working in the background: its subagents, the one waiting on your approval first with a Review button, and the commands and monitors the agent or a subagent left running, each with how long it has run, its latest output and a stop button, and a command started while the window was in the background shows as soon as you come back to it; stopping a subagent now stops the commands it left running too, and Claude Code chats show their background tasks in the same tray.
• Added argument hints to the message box, the Smart Bar and the Dynamic Island, as Claude Code shows them: typing a command and a space, such as /loop, /btw or /compact, or a skill whose SKILL.md has an argument-hint, shows what it takes dimmed after it until you type more.
• Added an Opacity slider to Settings › Chat Window, so you can make the Chat Window anywhere from clear glass to solid, previewing live as you drag; the Chat Window now starts at 75%, more solid than before.
• Changed the message box in a wide window to match the message list's width: the message box and the queued messages and cards above it now line up with the conversation instead of stretching across the window.
• Changed the Chat Window sidebar's chat rows: hovering a row now shows only Pin and Archive, resting on it shows its full title with its Project and Worktree beside the row, and right-clicking it opens a menu to rename, pin, mark as unread or read, move to a Project, fork, share, copy, open its folder in another app, archive or delete the chat.
• Changed a running chat in the Chat Window sidebar to show just its spinner, without the sheen on its title.
• Changed the chat list in the Chat Window to keep a chat's subagents folded under its row until you unfold them; the subagent you have open still shows, and a chat you unfold stays unfolded.
• Changed a Project's chats in the sidebar to show the first five, with Show more adding five at a time and Show less folding them back, remembered per Project; a chat that runs, waits for you, has a result you haven't read or is open still shows past them.
• Fixed Delete all chats in the sidebar's Chats menu deleting every chat, those in Projects and pinned ones included: it now deletes only the chats in the Chats list, says how many in its confirmation, and leaves chats that run or wait for you, or whose subagents still run, alone.
• Fixed Delete all chats in a Project's menu and in the Projects menu deleting chats that still run or wait for you, pinned chats, and Board task chats the list doesn't show: they now delete only the chats the Projects list, leave a chat whose subagent still runs, and say in your language how many they delete.
• Fixed Show more in the sidebar's Chats list folding back to 40 chats whenever a chat ran or got a new message.
• Fixed a chat that was never titled and began with a background task's notice showing the notice's raw tags as its title in the sidebar and in search.
• Fixed pinning, renaming or deleting a chat, and renaming or removing a Project, failing without a word: the sidebar now says when one fails.
• Fixed opening a chat from the chat list, a link in a chat, the archived chats list or a Board task so it comes in with its title, messages, queued messages, cards and the lines above them, instead of showing them a moment later and moving the conversation.
• Fixed a running chat opened from another chat or from a new chat showing only your message or an empty conversation until its reply caught up: it now opens on the reply written so far.
• Fixed the message box's model and reasoning, permission mode, coding agent mode and model, points pill and kept knowledge sources, and the header's agent name, changing a moment after a chat opened: each chat shows its own in the frame that opens it.
• Fixed the side panel's Files tab and the Worktree chip flashing empty or a no-folder notice before the chat's own files and Worktree showed when the chat opened.
• Fixed a Claude Code or Codex chat running on another Mac showing its message box before the read-only notice when opened again.
• Fixed the chat list redrawing the row of the chat being opened and the one open before it, which made the list flicker as you switched chats.
• Fixed the Chat Window not catching up when you switch back to it from another app: background commands started meanwhile now appear above the message box within a moment, and its other lists (Worktrees, artifacts, a Claude Code chat's state) read again too, instead of waiting until the chat runs again.
• Fixed the Changes review freezing the Chat Window for a second or more as it opened or showed again, and a file clicked in the chat while the side panel was closed now scrolls into view once the panel opens.
• Changed a turn's changed files under a reply to start folded.
• Changed an agent's Load Tools step to read as Looked up and the tools it loaded, and a run of steps that did other work no longer says it used a tool for it.
• Fixed a finished Install card, such as one saying a plugin still needs setup, disappearing into the folded list of tools the agent used.
• Fixed the / menu in the message box picking the command under a resting pointer instead of the best match when you press Enter.
• Fixed the terminal, the code editor in a file preview, the book reader and the workflow YAML editor showing their light colors on the Mac's dark glass.
• Fixed the Artifacts library listing only your 200 most recently changed artifacts: the next ones now load as you scroll toward the end.

Coding Agents

• Added /new and /clear to Codex chats and /clear (also found as /reset and /new) to Claude Code chats in the Chat Window, which start a new chat with the same agent in the same Project or Worktree, as each agent's terminal app does, and in Codex chats /new <name> or /clear <name> names the new chat and its Codex thread.
• Fixed the command menu of Codex and Claude Code chats opening again after a command with arguments was typed quickly, so Enter sends the command instead of picking it from the menu.
• Fixed Claude Code, Codex, Grok Build and Antigravity chats in the Chat Window keeping the stop button after a reply had finished.
• Fixed a Claude Code, Codex, Grok Build or Antigravity chat open in a window not showing a message sent to it from another window, the iPhone or a Hub task, or its reply, until the chat was opened again.
• Fixed question cards in Claude Code, Codex, Grok Build and Antigravity chats saying "Enconvo Agent asks"; they now name the agent that asks.
• Fixed a forked Grok Build or Antigravity chat, or a /btw Side chat of one, opening read-only with "This chat runs on another Mac. Continue it there." on the Mac it was made on; a new fork continues in the same folder with the same model and mode, and one made before this fix starts a new session that hears the conversation so far.
• Fixed a call about a Claude Code, Codex, Grok Build or Antigravity chat's waiting question opening a chat that showed "This chat runs on another Mac. Continue it there." instead of a box to type in; that chat is now one of your main Agent.
• Fixed a Claude Code or Codex chat that held messages before the agent started in it, such as one a Live Talk call began, opening read-only with "This chat runs on another Mac. Continue it there."; its first message now starts the agent with the conversation so far.
• Fixed editing a message in a Claude Code or Codex chat that held messages before the agent started in it: the agent keeps the earlier conversation, and Codex no longer keeps answering the message you replaced.
• Fixed forking a Claude Code or Codex chat that held messages before the agent started in it, such as one a Live Talk call began: a fork or Side chat no longer fails before the agent's first reply or from one of those messages, and a fork from a later message shows them too.

Claude Code

• Added Plugins to Claude Code's settings page: browse the installed plugins and every marketplace's, search them, install one, turn it on or off, or uninstall it, as /plugin does in the terminal, with a plugin listed in both the Anthropic Directory and an official Anthropic marketplace shown once, as "Anthropic Directory + official".
• Added /plugin to Claude Code chats: /plugin list shows the installed plugins as Claude Code's terminal does, including which ones need /reload-plugins, and the other forms point to the Plugins section of Claude Code's settings.
• Added /plugin install with a plugin's name to Claude Code chats, as the terminal does: it shows the plugin's details with Claude Code's choice of installing it for you, for everyone on the project, or for you in this project only, offers another scope for a plugin installed only for one project, asks to add the marketplace first with --marketplace <source>, then installs the plugin, turns it on in the chat, and asks for the options the plugin leaves unset before ending on "Installed and configured"; installing a plugin you already have opens the same form when options are still unset.
• Added /plugin enable, /plugin disable and /plugin uninstall with a plugin's name to Claude Code chats, which act at once as in Claude Code's terminal, asking first before an uninstall touches a project's shared settings or the plugin's saved data, and enabling a plugin asks for the options it leaves unset and ends on "Enabled and configured", or "Configuration skipped" when you close the form.
• Added /plugin configure <plugin> to Claude Code chats, which asks each of the plugin's options as a card and saves them as the terminal does, leaving sensitive options such as keys to the terminal so they never appear in the chat.
• Added /plugin marketplace list, add, remove and update to Claude Code chats, which list, add, remove and refresh plugin marketplaces at once as the terminal does, updating the plugins installed from a refreshed one.
• Added /plugin help and /plugin validate to Claude Code chats: help shows the terminal's plugin command list, and validate checks a plugin, marketplace or folder of skills, agents and commands and shows Claude Code's report.
• Added /reload-plugins to Claude Code chats, which applies plugin changes made since the chat started and lists what was loaded, and asks for --force when the reload would cost the conversation's cache, as Claude Code's terminal does.
• Added Claude Code's Rewind dialog to /rewind (/checkpoint, /undo) in Claude Code chats: pick an earlier message, then restore the code, the conversation or both to before it, as in the terminal, and a restored conversation puts the message back in the message box.
• Added the Background dialog to /tasks (/bashes) in Claude Code chats: see what runs in the background, open a shell's details with its command and latest output, and stop a task, as in the terminal; a shell that ended shows whether it completed, failed or was stopped, its real runtime and its output, and a background agent that just ended stays listed for 30 seconds, where picking it shows how it ran.
• Added the Context Usage card to /context in Claude Code chats: the colored grid of the context window, the model and tokens in use, usage by category, the auto-compact window, the MCP tools, agents, memory files and skills that fill it, and Claude Code's suggestions, as Claude Code draws them in the terminal, with /context all opening every list.
• Added Claude Code's /model dialog to Claude Code chats: /model lists Claude Code's models with the one in use ticked, and a model's card sets it as the default for new sessions, uses it for this session only, or adjusts its effort first, as in Claude Code's terminal app.
• Added Claude Code's /mcp dialog to Claude Code chats: /mcp lists the chat's MCP servers with their status and tools, and from a server's card you see its details and tools, reconnect, turn it off or on, sign in or clear a sign-in, as in Claude Code's terminal app; a message sent while the dialog shows runs once it closes.
• Added /resume to Claude Code chats in the Chat Window (/continue works too): it lists your saved Claude Code sessions from this chat's folder or every project, can show only the current branch's, searches their titles, and opens the one you pick as its chat, and /resume followed by a session ID or title opens that session directly, as Claude Code's terminal app does.
• Added /config (/settings) to Claude Code chats: every setting with its current value, then pick one and change it, as in the terminal.
• Added /plan to Claude Code chats: it turns plan mode on, sends a description to Claude in plan mode, or shows the current plan and opens it with /plan open, as in the terminal.
• Added /diff to Claude Code chats: list the folder's uncommitted changes, or a branch's changes against its base, and open any file's diff, as in the terminal.
• Added /copy to Claude Code chats, as in Claude Code's terminal: it offers Claude's last answer, or an earlier one with /copy N, whole or one of its code blocks or quotes, and puts the pick on the clipboard and in a file, and a new Skip the /copy Picker setting, also followed when it's on in Claude Code's own /config, copies the whole answer at once.
• Added /export to Claude Code chats, which copies the conversation or saves it to a file exactly as Claude Code's terminal writes it, asking how and for the filename unless one follows the command.
• Added /status to Claude Code chats, showing the same version, session, account, model, MCP server and settings details as Claude Code's Status tab.
• Added /permissions (and /allowed-tools) to Claude Code chats, which lists the allow, ask and deny rules and the workspace's folders and adds or deletes them, as Claude Code's terminal does.
• Added /add-dir to Claude Code chats, which adds a working directory for the session or remembers it in the project's local settings, as Claude Code's terminal does.
• Added /memory to Claude Code chats, which turns Auto-memory and Auto-dream on or off and opens your CLAUDE.md files and memory folders, as Claude Code's terminal does.
• Added /skills to Claude Code chats, which lists your skills with their source and size, searches and sorts them, and turns each one on, name-only, user-only or off for the folder, as Claude Code's terminal does.
• Added /hooks to Claude Code chats, which lists your hooks event by event and opens each one on what it runs and where it comes from, as Claude Code's terminal does.
• Added /sandbox to Claude Code chats, which picks the sandbox mode, the unsandboxed fallback and excluded commands for the commands Claude runs, as Claude Code's terminal does.
• Added /chrome to Claude Code chats, which shows Claude in Chrome's status and lets you install or reconnect the extension, pick the browser it uses, and set whether Claude Code turns it on by default, as Claude Code's terminal does.
• Added the Advisor picker to a bare /advisor in Claude Code chats: choose the advisor model or No advisor, with the current one checked, as in the terminal.
• Added the Auto-compact window dialog to a bare /autocompact in Claude Code chats: see the current window, then pick auto or 100k to 1M tokens, as in the terminal.
• Added /release-notes to Claude Code chats, which lists Claude Code's versions newest first and shows the notes of the one you pick, or of every version, as Claude Code's terminal does.
• Added /help to Claude Code chats: the General tab with its shortcuts, then Claude Code's default commands or your custom ones, as in the terminal.
• Added the folder's files to @ in Claude Code chats, as in Claude Code's terminal: @ lists the files and folders matching what follows it, a bare @ lists the folder's top level, and files the folder's .gitignore ignores stay out unless Claude Code's respectGitignore setting is off.
• Changed new Claude Code chats to start in Auto, where Claude decides which actions need your OK, unless Claude Code's settings name another default mode; a model without Auto starts in Accept edits.
• Changed Claude Code's tasks in a Project set to Approve for me to start in Claude Code's Auto mode, so they ask only before risky actions instead of stopping for approval on almost every command.
• Fixed a reasoning effort picked in a Claude Code chat showing as the model's default effort in later chats.
• Fixed Claude Code chats showing a loading line above an open card, such as /copy's, /model's or /mcp's: the card now takes its place, as in Claude Code's terminal.
• Fixed a ! command sent while Claude Code was still answering in a Claude Code chat leaving the chat stuck on Working after the command ran.

Codex

• Added Codex's /model dialog to Codex chats: it lists Codex's models with the one in use marked, then their reasoning levels with Max and Ultra under More reasoning…, and sets the pick as your default in Codex's config.toml or uses it for this session only, as in Codex's terminal app.
• Added Codex's /permissions to Codex chats: it lists Ask for approval, Approve for me, Full Access and Read Only as Codex's terminal app does, marks the current one and any your admin's requirements rule out, asks before Full Access, and keeps the pick for the chat, its forks and Side chats, also when it continues in a terminal.
• Added /resume to Codex chats in the Chat Window: it lists your saved Codex sessions from this chat's folder or every folder, sorted by last update or creation date and searchable, and opens the one you pick as its chat, as Codex's terminal app does.
• Added /delete to Codex chats in the Chat Window: after you confirm, it permanently deletes the session and its Codex thread, with its subagent threads, and opens a new Codex chat, as Codex's terminal app does.
• Added /import to Codex chats, which brings over your Claude Code or Cursor settings, instructions, MCP servers, hooks, skills, commands, plugins and recent chats into Codex as Codex's own /import does, with a summary to import everything found or choose items, and a report once the import finishes.
• Added /app to Codex chats on macOS, which continues the chat in the Codex Desktop app as Codex's own /app does, opening the same session there while the chat keeps it.
• Added /copy to Codex chats, as in Codex's terminal app: it offers the last answer whole, each code block and each quote in it, or right after /status the status and its fields, and puts the one picked on the clipboard.
• Added /cd and /rollout to Codex chats, as in Codex's terminal app: /cd followed by a folder moves the chat there, relative to its folder or from ~, and Codex works in that folder from the next message, while /rollout shows where the session's rollout file is.
• Added /mention to Codex chats, which writes @ in the message box to mention a file as Codex's own /mention does, whether picked from the / menu or sent.
• Added file search to the @ menu in Codex chats, which lists the chat folder's files and folders as Codex's own @ finds them, writing the picked path or attaching a picked image.
• Added /usage daily, /usage weekly and /usage cumulative to Codex chats, as in Codex's terminal app: the usage card's token activity opens on the calendar, a bar a week or the running total, and a switch on the card changes between them.
• Added Codex's Usage menu to /usage in Codex chats, which offers View analytics or Redeem reset as Codex's own /usage does, listing your usage limit resets and asking before using one.
• Fixed /model, /personality, /approve, /warnings and /archive sent while a Codex turn runs: they ask at once without holding up the message, and a pick made after the turn ends still takes effect and says what changed.

Grok Build and Antigravity

• Added Plan to the mode picker of Grok Build chats, also on Shift+Tab, and /plan and /view-plan (also /show-plan and /plan-view), as in Grok Build's terminal: Grok Build plans before it makes changes, /plan with a description starts planning that task, /view-plan shows the plan Grok Build wrote, and picking another mode takes it out of plan mode.
• Added steering to Grok Build chats: a message sent while Grok Build works goes into its running turn and shows in the chat at once, as Grok Build's terminal interject does, while Antigravity chats keep it for the next turn.
• Added /resume to Grok Build chats, as in Grok Build's terminal: it picks a session from the chat's folder to continue, narrows the list as you type, finds sessions by what was said in them under Extended search results, and /resume with a session id or title opens that session at once.
• Added /rewind (also /undo) to Grok Build chats, as in Grok Build's terminal: it lists the earlier turns, asks before going back, takes the conversation back to before the one you pick and puts that message back in the box to edit, and Grok Build's settings now have a Confirm Before Rewind switch.
• Added /new (also /clear) and /delete to Grok Build chats, as in Grok Build's terminal: /new starts a new chat, and /delete asks first, then deletes the session and its history for good and opens a new chat.
• Added /model (/m) and /effort to Grok Build chats, which work as in Grok Build's terminal: pick a model, its context window and its reasoning level from cards or type them, and the switch is saved as Grok Build's default for new sessions.
• Added /context-window to Grok Build chats, which works as in Grok Build's terminal: pick the current model's context window from a card or type its size, for the rest of the session.
• Added /context to Grok Build chats, which shows Grok Build's Context usage tab as a card, with the window's squares, the tokens of each part and when auto-compact comes, without sending the command to Grok.
• Added /session-info (also /status and /info) to Grok Build chats, which shows Grok Build's Session info tab as a card, with the version, session ID, folder, model, turn and context, where a click on a value copies it.
• Added /usage and /cost to Grok Build chats: they show your plan's usage limit with its bar and reset time, and the session's tokens, model calls and cost, as Grok Build's terminal does.
• Added /rename and /title to Grok Build chats, which name the session as Grok Build's terminal does, and Grok Build chats now take the titles Grok gives its sessions and pass a rename in Enconvo on to Grok.
• Added /copy and /export to Grok Build chats, as in Grok Build's terminal: /copy copies the latest answer, or the Nth latest with /copy N, to the clipboard or a new file, counting answers and naming a taken file the way the terminal does, and /export saves the whole conversation to a new file or copies it to the clipboard in the same Markdown.
• Changed Edit in Grok Build chats to keep the session, so Grok Build forgets the edited message and what followed and goes on with what came before.
• Fixed a Grok Build chat whose first message was a command the chat answers itself sending that command to Grok along with the next message.
• Fixed pictures Antigravity generates in a chat, which now show in the chat, also when an Antigravity session is opened again, and its permission cards and tool steps now name the tool instead of reading "Allow run …??" or staying on "Running …".

Loop

• Added starting /loop by asking the agent in plain words to do something repeatedly, such as checking something every few minutes or keeping an eye on it, which runs it in the chat as typing /loop would.
• Added /proactive as another name for /loop, as in Claude Code: typed in a chat it starts the same loop, and the message box's / menu finds Loop by that name.
• Changed /loop sent while a reply is still being written to wait in the queue and start once the reply ends, as in Claude Code, instead of being turned away; Cmd+Return queues it the same way.
• Changed a self-paced /loop to continue after Enconvo restarts: when you open its chat again, its next run comes back at the time it was set for, and if that time passed while Enconvo was closed or its runtime was restarting, the chat says the run won't happen and the loop continues when you reply asking it to.
• Changed the Smart Bar and the App Sidebar to show a waiting self-paced /loop above the input, with its next run and a stop button, as the Chat Window does.
• Changed stopping a reply so it also stops a self-paced /loop, as Esc does in Claude Code: the loop's next wake-up is canceled, including one the reply was setting as you stopped it, and the chat shows Pending wakeup cancelled, while a /loop with an interval keeps running.
• Changed a self-paced /loop's run to end with the delay the agent chose for the next run and its reason, such as Next run in 20 min and why, shown outside the run's folded work, as Claude Code prints them after each iteration.
• Changed when a self-paced /loop wakes up to match Claude Code: on the whole minute after the delay the agent picked, moved a minute earlier when that keeps a short wait under five minutes or a long one under an hour, and the agent's step and its list of scheduled tasks now show that time, with the wake-up listed as a daily one-shot task.
• Changed when a self-paced /loop wakes up on its own to match Claude Code: only a run the loop woke up for gets the fallback run about 20 minutes later when it sets no next run, not the turn you typed /loop in, and an agent that cancels the loop's pending wake-up by its ID during a run leaves the loop that fallback instead of ending it.
• Changed a self-paced /loop to pick its wait with the model's prompt cache in mind, as Claude Code does: the agent checks again within about 4 minutes while watching something closely, or waits 20 minutes or more, so fewer runs read the whole chat again uncached.
• Changed a self-paced /loop to tell the agent how many runs in a row found nothing to do, as Claude Code does, so it can ease off or notify you instead of reporting the same quiet check again.
• Changed a self-paced /loop typed again with the same prompt to keep its 7-day limit, as Claude Code does, instead of starting a new 7 days each time; a stop, a new prompt or an hour without a run starts them anew.
• Changed a self-paced /loop to schedule its next run without a Load Tools step first, as Claude Code keeps its wake-up tool always at hand.
• Changed a /loop with no prompt to give the agent its loop.md, or the built-in maintenance prompt, once, as Claude Code does: later runs point back to it until the file changes, so a long loop.md no longer fills the chat's context again with every run.
• Changed a /loop with no prompt to cut a loop.md longer than 25,000 characters at a line break and tell the agent it was cut, as Claude Code does, instead of cutting it off mid-line without a word.
• Changed a /loop with no prompt to notify you mid-loop as Claude Code does: once when it can't go on without you, finds nothing to do three runs in a row, ends, or sees a major update such as a failed CI (continuous integration) run, not on every run.
• Changed a /loop with no prompt, such as /loop 2h, to leave out the question about making an Automation and the line saying the loop stops when Enconvo quits, as Claude Code does, and to say in its confirmation whether it runs your loop.md or the built-in maintenance prompt.
• Changed when /loop's scheduled tasks run to match Claude Code: a task every five minutes first runs 4 minutes 45 seconds after it is made, even when the agent schedules it on its own, a daily task usually runs 30 minutes after its time, and a run held up by a busy chat moves the next one to after it.
• Changed /loop's scheduled tasks to read cron expressions as Claude Code does: a day field that covers every day, such as 1-31, leaves the date to the other field, a step on a single number such as 5/15 is refused with the range to write instead, and a task's schedule reads as Every 5 minutes, Weekdays at 9:00 AM or Every Monday at 9:00 AM.
• Changed /loop's steps in the chat to read as Claude Code's: scheduling a task shows Scheduled with the task's ID, its schedule and next run, listing the tasks shows each one with its next run, and canceling one shows Cancelled with its ID.
• Changed /loop's list of scheduled tasks in the chat to read as Claude Code's: each task is one line with its ID in bold, then its schedule, next run and prompt, marked recurring or one-shot and session-only, a one-shot task shows its schedule, an empty list says No scheduled jobs, next run times are rounded to the minute with the weekday on other days, and a long prompt is cut by its width on screen, so one in Chinese or with emoji is cut where Claude Code cuts it.
• Changed /loop's lines for scheduling, listing and canceling tasks to stay in view after a turn ends, outside its Worked for line, with the listed tasks shown under their line without opening it, as Claude Code's terminal shows the rows under these calls.
• Changed what /loop's scheduling tools tell the agent to Claude Code's own wording, so the agent reports each task's schedule, whether it repeats, and its next run to the minute as Claude Code does.
• Added a way for agents to list the commands their chat started that are still running, so a self-paced /loop checks whether its monitor still runs before starting another and finds it to stop when the loop ends, as Claude Code does with its task list.
• Fixed an agent that waited for something with a long-running command, such as a self-paced /loop watching for a file to appear, not hearing when the command ended: it now continues as soon as the command exits, as with background commands, and also hears when a dev server it started stops on its own.
• Fixed a self-paced /loop's watch for an event running on with no time limit: it now ends after 30 minutes, as Claude Code's does, and wakes the loop when it ends.
• Fixed Esc in the Smart Bar, the App Sidebar and the Dynamic Island closing or folding it while a self-paced /loop waited for its next run: Esc now stops the loop and shows Pending wakeup cancelled, as in the Chat Window and Claude Code.
• Fixed a self-paced /loop stopping without a word when Enconvo's runtime restarted during one of its runs: the loop now goes on from its fallback run about 20 minutes later, keeping its run count.
• Fixed a /loop task that came due while a reply was running being dropped when Enconvo's runtime restarted before the reply ended: a one-time task now still runs once the chat is free, and a repeating task runs the time it was waiting for.
• Fixed a self-paced /loop whose wake-up runs a prompt without /loop going quiet when a run neither schedules the next one nor stops: it now gets one fallback run about 20 minutes later, as Claude Code's does.
• Fixed a self-paced /loop that reached its 7-day limit telling the agent the loop was stopped, so the agent now says the loop expired and that /loop starts it again.
• Fixed a /loop task or wake-up canceled at the moment it came due still running once.
• Fixed /loop's list of scheduled tasks showing a next run time already gone by for a task that came due while the chat was busy; it now says the task runs now, as Claude Code's list does.
• Fixed /loop one-time tasks set inside the hour the clock turns back, which now run that night as in Claude Code instead of a year later, and changed the errors for a date that never comes and for too many tasks to Claude Code's wording.
• Fixed /loop asking whether to make an Automation when the prompt only mentions something nightly or daily, such as a nightly build, instead of asking for a daily schedule; the agent now asks only when the words say how often the loop runs, as Claude Code does.
• Fixed /loop with an interval of 45 to 59 minutes, which runs every hour, leaving out the line under its confirmation that the loop stops when Enconvo quits.
• Fixed /loop every five minutes, or an interval alone ending in a full stop, running the loop's default tasks every five minutes; as in Claude Code, it now shows how to use /loop and starts nothing.
• Fixed a self-paced /loop refusing a wake-up whose reason is empty, and telling the agent its wait was cut short when it only asked for a fraction of a second, as Claude Code does neither.
• Fixed a self-paced /loop whose prompt has &, < or > changing it on every run, such as && turning into &amp;&amp;, which also kept the loop from ever reaching its 7-day limit.
• Fixed a /loop sent during a reply starting anyway when you stopped the reply just as it finished, such as with Esc in the Smart Bar or Stop on the phone; it now stays in the queue for you to send, as the other queued messages do.
• Fixed a /loop waiting in the queue moving behind the messages sent after it when it couldn't start at once, so those ran before it.
• Fixed a Loop's runs, and other turns that start on their own, not showing in an open chat: each run now appears with its line and reply as it happens, instead of taking the place of the reply before it.

Side Panel Doodle

• Fixed the side panel Doodle quoting a word split across styles, such as a token of highlighted code, a highlighted search term or a bold first letter, with spaces inside it, or cutting part of it off at the ends of a passage dragged with the Paragraph tool.
• Fixed the side panel Doodle's Paragraph tool running words together in Arabic or Hebrew text when they sat in links or styled words of their own.
• Fixed a Doodle mark over Chinese or Japanese text in the side panel quoting it with a stray space, splitting a word, where a line breaks before a link or a bold word.
• Fixed the side panel Doodle sending a broken character when it shortened the words under a mark right inside an emoji.
• Fixed a Doodle mark on a file or web page in the side panel quoting words you can't see under it, such as code or a side list scrolled out of sight, or the end of a title cut short.
• Fixed a Doodle mark on a web page in the side panel quoting words hidden under a bar fixed at the top of the page or a dialog over it.
• Fixed the side panel Doodle's Comment tool quoting the words of hidden parts of a page, such as a closed menu, with the paragraph or passage picked, and refusing a passage dragged across them.
• Fixed a region boxed with the side panel Doodle's Comment tool on a web page quoting only its first 300 characters.
• Fixed the side panel Doodle's Comment tool picking a whole list on a web page laid out in a table, such as a news site's front page: a table or code taller than most of the view is now read paragraph by paragraph.
• Fixed the side panel Doodle's Comment tool taking a table, code block or quote set in a figure (as WordPress and Jekyll sites set them) for a picture with no words: it now picks them as what they are.
• Fixed the side panel Doodle quoting nothing for a picture whose alternative text is empty: it now quotes the picture's caption, label or title.
• Fixed the Doodle Paragraph tool in the side panel picking a long run of lines in a log or a long code file, far from the line you point at, where no blank line comes near; it now picks that line.
• Fixed the side panel Doodle reading only the word at one end of a line: a line now quotes the words it underlines or strikes through, whichever way it was drawn.
• Fixed the side panel Doodle sending a picked code block or table to the AI as one line: it keeps its lines, and code keeps its indents.
• Fixed the side panel Doodle's Screenshot tool taking in the chat or window beside the side panel when its box was dragged past the panel's edge.
• Fixed side panel Doodle marks on a long web page or document being slow to read the words under them, or losing those words, when many were moved or pasted at once, and a mark far down a long paragraph quoting nothing.
• Fixed scrolling a side panel tab while Doodle is on: a slight sideways drift over a wide code block or table no longer stops the page from scrolling, and content that can't scroll sideways no longer slides out of place.
• Fixed a wide table or code block on a right-to-left page, such as one in Arabic or Hebrew, not scrolling sideways, and vertical text, such as Japanese set in columns, not scrolling on with the wheel or trackpad, in a side panel tab while Doodle is on.
• Fixed ⌘C and ⌘X not copying the text selected on a web page in the side panel while Doodle was on and no mark was picked.
• Fixed the side panel Doodle's select tool and eraser missing a long, flat ring, such as one drawn round a line of words, near its ends, and a press on a wide arrow's head away from its shaft.
• Fixed cutting a mark in the side panel Doodle and pasting it, which lost the mark's comment: a cut and paste now moves the mark with its comment, a copy still starts without one, and a cut brought back with Undo no longer takes its comment along when pasted in another chat.
• Fixed Copy view as image in the side panel Doodle, which failed in the Mac app: the copy now holds the newest marks, and it can be used as soon as a view has marks.
• Fixed the Doodle list above the message box losing a note's edited words when the pencil or a click outside the list ended the edit.

Memory

• Fixed a Memory about you, such as a food allergy, that you mentioned in a Project's chat being kept for that Project only, so your other chats never used it; agents and Memory Review now save such Memories for every chat and keep only the Project's own facts in the Project.
• Fixed an agent outside a Project reaching a Project's Memories: being shown them when it saved one much like them, which led it to delete what you had just asked it to remember, deleting or bringing them up when you asked it to forget or change something, such as a Project's note about a former colleague when you asked to forget your old job, and finding them when it searched your Memories and past chats together or read a person or topic from your Memory's Concepts.
• Fixed a subagent started in a Project's chat not knowing the Memories saved for that Project, and keeping what it saved out of the Project.
• Fixed agents sometimes describing a Memory about you as someone else's, such as calling your own food dislikes your child's while planning her party.
• Fixed agents applying a Memory about someone else to a task that is only yours, such as holding a lunch you order for yourself to your partner's food allergy.
• Fixed the agent sometimes saving a change, such as a new home address, as a second Memory beside the old one once you have many Memories; it now finds the old Memory and updates it.
• Fixed Memory Tidy deleting an older Memory when a later one changed only one detail in it, which lost the rest, such as where your router is after your Wi-Fi password changed; the two are now merged.
• Fixed agents deleting a whole Memory when you asked them to forget only one detail in it, such as removing your Wi-Fi network's name along with its password.
• Fixed agents writing the mistake and its correction into a Memory when you corrected part of it, such as noting that a mango allergy had been saved by error, instead of keeping only what is true.
• Fixed the agent adding things you didn't say to a Memory, such as the day it was saved, a date it guessed, a detail from another Memory or a remark of its own; it now saves only what you said.
• Fixed Memory Review adding a detail you didn't give to a Memory it learns from a chat, taken from another Memory, such as calling a friend you named only by first name your manager because your manager has that name, or adding the city you live in to a place you mentioned.
• Fixed agents and Memory Tidy making up a day for a Memory that gives only a month, such as a child's birthday month written with the age right after it: Memory Tidy keeps the month, and agents plan around the day and ask which day it is.
• Fixed agents refusing to remember a password or door code you explicitly asked them to keep; it stays hidden from other chats until you ask for it, and a password written without "is" or a colon after the word "password" is now hidden too.
• Fixed a Memory holding a saved password hiding the words after the password from agents too, which made them say they could not confirm a changed password was saved.
• Fixed Memory Tidy sometimes rewriting or deleting a Memory that holds a password or key you asked to keep, which lost the secret.
• Fixed agents missing details inside a file or web page you saved to Memory, such as a rule in a team handbook, and looking it up on the web instead.
• Fixed agents asked to keep a file sometimes saving only a note about it, because adding the file to a knowledge base failed or Memory kept only their summary of it.
• Fixed agents asked to replace a file saved to Memory with a new version sometimes overwriting your original file on disk, which left the old version in Memory.
• Fixed asking a voice call to note something down for later, such as where you parked, saving it as a reminder or a note instead of a Memory, so a later call asking for it could not find it.
• Fixed the agent in a Telegram or Discord group sometimes saying it had remembered something a member asked it to keep, when Memory is off in group chats; it now says Memory works only in a private chat with it.
• Fixed an open chat changing a Memory back, or asking which was right, after you edited that Memory on the Memory page; the agent now goes by your edit.
• Fixed a Memory the agent saves during a chat having no link back to that chat on the Memory page: the link shows the chat's current name, points to the chat where the Memory last changed, and goes away when the chat is deleted.
• Fixed the agent not knowing when or where it learned something it remembers; asked, it now gives the day the Memory was saved and the chat's name, or says that chat was deleted.
• Fixed Undo after forgetting a Memory on the Memory page, Save again on a chat's Memory card, and Undo or Redo in the Memory history bringing a deleted Memory back as a new one for every chat: it now returns to its Project and the Agent that saved it, with its original dates.
• Fixed agents bringing back a Memory you deleted when asked what they remember about you, because the chat where you first asked them to remember it was still in Chat History, and mentioning that you had asked them to forget something.

Workflows

• Added who does each step to the workflow editor: every step card names its Agent, Member slot or tier, with the model and reasoning effort it runs on, and the inspector lets you pick the Agent, slot or tier template, a model and a reasoning effort for that step's run only.
• Added a reasoning effort (low, medium or high) to workflow prompt and agent steps, for that step's run only and settable from the run's inputs; an agent step can now pick its Agent from the run's inputs, falling back to the run's default Agent when its pick names none, and a workflow with Member slots can mix them with named Agents and its own models.
• Changed the workflow list so every row is a workflow, with no separate Playbook kind: a row shows its steps, parallel lanes, schedule and when it was edited, a Member slots · N tag when it declares slots, a Pinned in steps · N tag for steps that name their own Agent or model, and a Used in column with how many Projects use it; the list filters by All, With Member slots and With a schedule.
• Changed the workflow editor, the Board and the Run log to say workflow and Member slot where they said Playbook.
• Changed where a Project's Member slots are filled: open a workflow in a Project, from its title bar, the Used in column or a Board task's Open the workflow, and a bar above the canvas shows each slot with the member filling it there, the slots the lead filled from the Project's members, or that the Project has no members yet with Add members; a slot's chip, or its badge on a step card, opens a picker with the slot's description and the lead's suggestions, the change applies to that Project only, Run waits until every slot is filled, and Project settings no longer have their own Playbooks page.
• Changed Member slots so a Project's Agent fills a workflow's empty slots with the members whose Member roles match them, and a slot you fill or empty yourself stays as you left it.
• Changed workflows with Member slots so any of them can be given to a Board task in a Project as soon as its slots are filled there, without adding it to the Project first, and one that can't run yet says whether the Project needs members or which slots are still empty.
• Changed Board tasks that follow a workflow to run steps that name their own Agent through the Board, like slot steps: in the task's worktree, at its Permission mode and within the Project's Boundary, each shown on the task's card with its Agent, and failing with the reason when that Agent doesn't exist, is paused or isn't signed in, or when Enconvo can't read the list of Agents just then; a step that names a Coding agent by its own key, such as coding_agent|codex, runs on the Agent that wraps it.
• Fixed the workflow editor's step palette and Add step list leaving out the Agent step in a workflow with Member slots, though such a step may name its own Agent: you no longer have to write it in YAML.
• Fixed Workflow agent steps whose agent: template picked an Agent that doesn't exist failing instead of running on the default Agent; an Agent key written out still fails as written.
• Fixed Workflow steps whose AI request failed (the provider refused the model or answered with an error) or couldn't start (the provider couldn't load) finishing as done with the error as their reply; the step now fails with that error.
• Fixed a Board task's later-slot workflow steps working outside the Project's Boundary and losing the task's worktree when the Project's folders left it out.

Dynamic Island and File Search

• Added date groups, scrolling through every chat, a title search and pinning to the Dynamic Island's History, which showed only about five chats before, with a button at the bottom that opens all chats in the Chat Window, or searches their messages there for what you typed.
• Fixed the Dynamic Island's History keeping the Mac busy while one of its chats was running.
• Fixed chat previews in the Dynamic Island showing the model's thinking or Markdown marks instead of its reply.
• Added a Search Folders setting to File Search, with a My Folders choice in Search In that looks only in those folders, Desktop, Documents and Downloads unless changed, one folder per line.
• Fixed File Search treating a path such as ~/.claude or /Library/Logs/x as a loose name search, which showed an unrelated folder whose name only resembled the letters typed while the hidden folder named stayed out: a path now looks in the folder it names, whatever the Search In scope, for the name after its last slash, with the exact path first and that folder's own files next, finds a hidden file or folder the path names, and a path ending in a slash, such as ~/ or ~/.claude/, lists the folder.
• Fixed File Search treating a Search Folder inside Library or a hidden folder, such as iCloud Drive or ~/.config, as noise: its files now show and rank by name like any other folder's, with only the hidden and cache folders below it still set aside.
• Changed File Search to list an app first when you type its name, a typo of it or a word of it, such as Safari, "safair", or "chrome" for Google Chrome, before the files named after it, which used to come first when they were newer, while an app that only contains the letters, such as Xcode for "code", still comes after a file named as typed.
• Changed File Search to show a hidden file named as typed with a dot in front, such as .zshrc for "zshrc", as the exact match, while other hidden files still need the dot typed or the Show Hidden Files setting; the Finder now comes first for "Finder".
• Changed File Search's index to skip file contents, since it searches names only, so it reads no files after a change and keeps no content index on disk.
• Fixed File Search waiting while its index was busy applying a burst of disk changes: a search it doesn't answer within 1.5 seconds now falls back to Spotlight.
• Fixed File Search without Full Disk Access never showing some Desktop, Documents and Downloads files on Macs where Spotlight takes over three seconds to list them: the listing now runs to the end, and typing on narrows it instead of asking again.
• Changed File Search to leave one or two typed letters to the index alone, since Spotlight listing those takes seconds and shows nothing useful.

Skills, Plugins and Sync

• Added Add from GitHub to the Skills page: paste a link to a repository with SKILL.md at its root, or to a skill's folder inside one, and the skill installs into your skills folder and opens, with Enconvo remembering where it came from.
• Added Check for updates to the Skills page for skills added from GitHub: each skill whose repository has moved on gets an Update available mark and an Update button in its details, and the result offers Update all.
• Fixed the launch-time update of a skill added from GitHub into a folder other than ~/.enconvo/skills, which installed the new copy into ~/.enconvo/skills instead of the folder the skill was in.
• Added Skills and Plugins to Desktop Sync, each with its own switch under Settings › Sync: the skills you install or make, with whether each is on, and the plugins you install from the Enconvo store, GitHub, npm or a link now sync between your computers, which install them from where they came from and uninstall what you uninstall, while plugins that come with Enconvo, plugins you are developing, Connectors and sign-ins stay on each computer.
• Removed the Bland Connector from the Plugins page; if you already installed it, it stays installed.
• Fixed a Connector, such as Twilio, missing from the Plugins page on a Mac where its plugin folder had been left with only empty folders in it.

Models, Sign-in and Settings

• Added OpenCode and OpenCode Go as AI model providers: add your OpenCode Console API key once under Settings › AI Model to chat with OpenCode's pay-as-you-go models, such as Claude, GPT and Gemini, or with the models of your OpenCode Go subscription.
• Fixed the model picker listing a provider you had only looked at in Settings: a built-in provider such as OpenRouter or DeepSeek now appears in the pickers once its API key or sign-in is saved, so opening its page no longer adds it with no key, while Enconvo Cloud, local models and providers you created stay listed as before.
• Added a Back to Enconvo button to the page an OpenAI, xAI or Antigravity sign-in ends on in the browser, and a sign-in started in Settings or the guide now brings that window back by itself once it connects, as one started from a chat does.
• Changed signing in with a Claude subscription or Claude Code: copying the code the sign-in page shows is enough, and Enconvo signs in without it being pasted; the code stays out of Clipboard History.
• Fixed AI Commands that follow the default model, such as PopBar buttons, the Translator and Dictation's Polish and Translate, showing your default model in their model button and picker while running on the All Commands model, so an Out of Cloud points notice could name a model the button never showed; the button now shows the All Commands model, the picker's Use default reads All AI Commands for these commands, a reasoning effort picked for one of them now applies, and the notice after picking another model points to Settings › AI Commands.
• Added Upgrade to Pro and Upgrade to Teams in Settings › Account and the Plans page for Plus and Pro license owners: what you paid for your license counts toward the new one, so you pay only the difference, promo codes still apply, every Mac you use keeps a seat on Teams, and a refund within 30 days gives your old license back.
• Fixed Automation showing as Not granted on the Permissions page and in the Computer Use permission prompt after a relaunch, when the grant stood but System Events wasn't running.
• Fixed Set up Telnyx in Settings › Agent Phone doing nothing when Telnyx was already connected in Plugins: it now waits while the Telnyx plugin updates, and says so when Agent Phone can't use it yet.

Voice

• Fixed a reply being read aloud talking over Dictation, as when you clicked the mic in the Dynamic Island's, Chat Window's or Smart Bar's input or pressed the Dictation hotkey while the answer was read: the reading now pauses before Dictation starts listening, including the sentences still to come, and reads on from where it paused when you finish or cancel, while a reading you had paused yourself stays paused, one that was stopped or replaced while you dictated doesn't come back, and Voice Command keeps its own behavior.
• Changed voice calls on the Mac to pause her the moment you start talking over her, before your words are recognized: your words then stop her for good, while an "mm", a cough or a laugh lets her go on from just before where she stopped, and talking while your microphone is muted doesn't pause her.
• Fixed a voice call that hears you with a local speech model being slow to stop when you talk over her: it now listens again about every quarter second while you go on, where it waited a second and a half after first hearing too little, and a word or two said over her, such as "wait", stops her as soon as you pause after it.
• Fixed a voice call on the Mac or iPhone staying on after you said goodbye in a longer sentence, such as "I just need to say hi and say goodbye, bye": Jev now reads whether your words and her reply end the call, in any language, and the call hangs up once her goodbye has played; a goodbye to someone else, or more words after hers, keeps the call on.

iPad and Mac Screen

• Added iPad support to the iPhone app: the chat list stays beside the chat on a wide screen, the side panel opens beside the chat with Full Screen in its tab strip, chats keep to a readable column, the app works in a resizable window, and strings that named the iPhone name the iPad.
• Added the iPad menu bar's commands with the Mac's shortcuts: File › New Chat (⌘N), and in View Show Sidebar (⌃⌘S), Show Side Panel, Full Screen Panel, Doodle, a Mac chat's Files (⌘P), Changes (⌃⇧G), Terminal (⌃`) and Side chat (⌥⌘S) in the side panel, Open Location (⌘L), Reload Page (⌘R), Back (⌘[) and Forward (⌘]) for the side panel's page, Show Next and Previous Tab (⌃⇥ and ⌃⇧⇥), Show Tab (⌘1 to ⌘9) and Reopen Closed Tab (⇧⌘T); Edit › Find (⌘F, ⌘G and ⇧⌘G) finds in the side panel's page.
• Added a divider between the chat and the side panel on iPad that you drag to resize them, as on the Mac; the width you leave stays for every chat.
• Added side panel windows on iPad: drag a web page's tab out of the side panel, or choose Open in New Window from its menu, and it opens in a window of its own with its tabs and New Tab, a button that brings back the chat it came from, and closing its last tab closes it.
• Added Apple Pencil to the side panel's Doodle on iPad: once the Pencil marks, a finger scrolls the page, a Draw with Finger switch in Doodle's bar lets a finger mark too, and a hovering Pencil shows where it will land.
• Added the Live Talk avatar beside the chat on iPad, as on iPhone Duo's inner display, with Full Screen and Back to the chat in her header; a link or the side panel opened from the chat meanwhile shows over the chat.
• Changed notes at the top of the iPad app to show as a centred line instead of a bar across the whole window.
• Added Mac Screen to the iPhone and iPad app: see your Mac's screen live from Remote, take control to click, scroll and type on it with trackpad or direct touch gestures and the phone's keyboard, with the toolbar on iPhone Duo's bar edge and a trackpad half when half folded; agents using the screen wait while you have it, a dropped connection, such as the phone moving to another network, connects again by itself, and Settings › Connections › Screen on the Mac turns it off.
• Added a Screen button to your connected Mac in a new chat's Runs on menu in the iPhone and iPad app, with a one-time tip that points to it, so Mac Screen is a tap away when you start a chat; each Mac there now says only whether it's connected.

iPhone and Android

• Added the / menu to Enconvo chats in the iPhone and Android apps: typing / at the start of a message lists /loop, /compact and your skills, with what each takes, and typing a command and a space shows its argument hint dimmed after it, in Claude Code and Codex chats too.
• Added interactive apps to Mac chats on the iPhone, where they run as they do on the Mac, Fix button included.
• Changed the Work plan card in the iPhone app to say what each task waits for, as on the Mac: before Start, the tasks it runs after or its turn in a folder the tasks share, and once the plan runs, the task, the lead's check or the free slot it waits for.
• Changed the message box of a new chat that runs on your Mac in the iPhone and iPad app to match the Mac chat's, with the text above a row of round buttons.
• Changed the iPhone app's message box so Return on a hardware keyboard sends and Shift-Return starts a new line, as on the Mac.
• Changed the side panel's Terminal tab to show only the Escape key above the keyboard while a hardware keyboard is attached, on iPad and iPhone.
• Added Android phones to the phone apps Mac agents can use: an agent can ask a paired Android phone where you are or set an alarm on it, as it can an iPhone.
• Fixed web pages at plain http:// addresses not opening in the iPhone app's side panel or from a link in a chat: they now load as they do in Safari.
• Fixed the iPhone app offering Move to Project for a chat that works in a worktree, where moving always failed.
• Fixed sliders, drawing areas and sideways-scrolling parts of chat widgets on the iPhone, where a drag on them scrolled the chat or opened the sidebar instead.
• Fixed questions the Agent asks in a chat started on the iPhone, which were cancelled when the Dynamic Island on the Mac closed while you were still answering them on the phone.
• Fixed a background command's output in a Claude Code chat on the iPhone and iPad app staying on its first lines on iOS 26 instead of following the newest ones.
• Fixed the Remote row at the top of the iPhone app's chat list cutting off whether your Mac is online when its name is long.
• Added xAI on your own account to the iPhone and iPad app's Text-to-Speech, Dictation and Live Talk: with an xAI API key or the Grok sign-in in Settings › AI Model › xAI, answers read aloud in xAI's voices, dictation streams to xAI's real-time speech recognition, and calls in phone chats hear and speak with xAI, all straight to xAI with no Cloud points.
• Added xAI on Enconvo Cloud to Dictation and Live Talk in the iPhone, iPad and Android apps, paid for with Cloud points and needing no xAI account, beside xAI on your own account.
• Changed the Text-to-Speech, Dictation and Live Talk settings in the iPhone, iPad and Android apps to list their services the way Settings › AI Model lists its providers, with Default on the one in use and a Cloud tag on each Enconvo Cloud service, as on the Mac; Text-to-Speech lists xAI on Enconvo Cloud and xAI on your own account as two services, and each service opens its own page with the account it uses, the model to download, the voice with a sample, and Set as Default.
• Fixed Live Talk calls in Mac chats in the iPhone, iPad and Android apps with Listen or Speak set to xAI on your own account: they now use the xAI account on the phone, the Grok sign-in included, so a call no longer ends with an xAI sign-in error when the Mac has no xAI account, and replies are spoken in the xAI voice you chose instead of a Gemini voice.
• Changed Kokoro voices in the iPhone app to work on iOS 27.2 and later too, whose Core ML no longer crashes them during long readings; they stay unavailable on iOS 26.4 to 27.1.
• Fixed voice samples in the iPhone and Android apps' Text-to-Speech and Live Talk settings cutting into a voice chat in progress; they now ask you to end the call first.
• Fixed Text-to-Speech settings in the iPhone and Android apps showing a system voice that's no longer on the phone by its identifier instead of its name.
• Fixed the iPhone and Android apps naming the regions of system voices in the phone's language instead of the app's.
• Fixed the voice filters in the iPhone and Android apps hiding the filter you pick at the screen's edge.
• Fixed Dictation's Languages and Details in the iPhone app treating Cantonese as Chinese for models other than Soniox.
• Fixed the Dictation row in the iPhone and Android apps' Settings saying the model isn't downloaded while it downloads.
• Fixed the Android app's Text-to-Speech settings showing voices you install in Android's settings only after the app restarts.
• Fixed the Android app cutting off a voice sample you start right after leaving a voices page.
• Fixed the Android app letting go of the dictation model after Mac Screen or another screen opened over it, so the next dictation waited for the model to load again.
• Fixed the Android app's Settings opening a page twice, or going back two pages, after a quick second tap.
• Fixed the Android app sometimes going back two pages when a change from your Mac arrived just as you left a voices page.
• Fixed the Android app sometimes closing when a web page or Terminal in a chat's side panel stopped, such as after the app was away for a while; the tab now loads again on its own.

Answers on the iPhone and Android

• Added charts to answers in the iPhone and Android apps: the bar, line, area, pie, donut and funnel charts the Agent draws now show on the phone, where they were left out before.
• Added sliders, number fields, pickers and switches to answers in the iPhone and Android apps: a calculator or another view the Agent builds with them works its results out again as you change them, as on the Mac, where the inputs were left out and the results blank before; a button can send your values along with its prompt, and Reset Values in the card's menu puts them back.
• Added quizzes you can answer in the iPhone and Android apps: pick an option and check it to see whether it's right, the answer and why, then try again, as on the Mac, where the phone showed only the question and its options before.
• Added flashcards you can study in the iPhone and Android apps: tap a card or Flip card to see its back and note, and move through the deck with the arrows, as on the Mac, where the phone showed the fronts and backs as a list before.
• Added checklists you can tick off in the iPhone and Android apps: tap an item to tick it or clear it, with the count and a progress bar filling as you go, as on the Mac, where the phone showed the items as a list before.
• Added tabs in the iPhone and Android apps, as on the Mac: a plan in parts, such as the days of a workout, shows one part at a time under a row of tabs you tap, where the phone showed every part in a long list before.
• Added step-by-step walkthroughs in the iPhone and Android apps, as on the Mac: an explanation in stages, such as how a recipe or an algorithm goes, shows one step at a time with Previous, Next and a bar of the steps you can tap, where the phone showed every step in a list before.
• Added schedules to answers in the iPhone and Android apps: a day or week the Agent plans shows as an agenda by day, each block's time beside a bar of its kind's color, with a legend of the kinds and their hours, as on the Mac, where it was a plain list before.
• Changed option comparisons in the iPhone and Android apps to look like the Mac's: plans or products to choose from show as cards, each with its price, what it includes and lacks, and the best fit ringed with its badge, where the phone showed a plain list before.
• Changed timelines in the iPhone and Android apps to look like the Mac's: itineraries, histories and the steps of a process show each event as a dot on a line, its time beside its title and its details under them, where the phone showed a plain bulleted list before.
• Changed flowcharts in the iPhone and Android apps to look like the Mac's: steps go down a column with arrows between them, a start or end in a small capsule, a decision in a diamond, and its Yes and No branches side by side under their labels, where the phone showed a numbered list before.
• Changed roadmaps in the iPhone and Android apps to look like the Mac's: phases stand side by side, each a colored header over its progress bar with its title and percentage, and the overall completion shows under them, where the phone showed a bulleted list before.
• Changed trees in the iPhone and Android apps to look like the Mac's: org charts and breakdowns show as colored boxes joined by lines, or as an indented outline when they are too wide for the screen, where the phone showed a bulleted list before.
• Changed cycles in the iPhone and Android apps to look like the Mac's: stages sit around a loop as numbered colored circles joined by arrows, with the title in the middle and a key of what each stage does under it, where the phone showed a numbered list before.
• Changed maps in the iPhone app to look like the Mac's: places are marked by number on an Apple Maps picture, joined by a dashed line when they make a route, with the numbered list of them under it, where the phone showed a plain list before; the Android app shows the numbered list.
• Changed weather cards in answers in the iPhone and Android apps to show like the Mac's: the conditions now with their tiles, the next six hours and the days ahead, or the one-row card, loaded live for the named place or for where the phone is, where the phone left the weather out before.
• Changed flow charts and trees in answers in the iPhone and Android apps to show their icons as on the Mac: each box's icon sits above its name, and in an outline it takes the place of the dot, where the phone left the icons out before.
• Changed step lists in answers in the iPhone and Android apps to show as on the Mac: each step is a row with its icon in a tinted tile beside its name and description, where the phone showed a plain bulleted list before.
• Changed colored words in answers in the iPhone and Android apps to show as on the Mac: right answers in green, mistakes in red and other colored or bold words keep their color and weight and stay on the line with the words around them, where the phone showed them plain before.
• Changed progress bars in answers in the iPhone and Android apps to show as on the Mac: each one is a thin bar in its color under its label and share, with its note under it, and one past its limit, such as spending over a budget, marks where the limit falls and stripes the part beyond it, where the phone showed a row of number tiles before.
• Changed Mermaid diagrams in answers in the iPhone and Android apps, such as sequence and state diagrams, to show drawn as on the Mac while the Mac is connected, where the phone showed their code before; a tap opens the diagram to zoom in.
• Changed the main next step under an answer in the iPhone and Android apps to look as on the Mac: a button the Agent makes full width spans the card on a line of its own as a rounded pill with an arrow after its label.
• Fixed an answer in the iPhone and Android apps losing everything after a part that changes with a choice, when that part's text held an apostrophe.

Languages

• Fixed Chinese switching between the formal "您" and the informal "你", as an empty chat asking "有什么可以帮您？" and the Mac's location and photo permission prompts; it says "你" throughout.
• Fixed Traditional Chinese mixing in mainland wording, as an agent called 智能體, 智慧體 or 智慧代理 on different screens, and 保存, 點擊 and 剪貼板 for save, click and clipboard; it says 代理程式, 儲存, 按一下 and 剪貼簿 throughout.
• Fixed Korean switching from the formal "합니다" to the casual "해요" on some screens, as a stopped task saying "작업을 멈췄어요", and Simplified Chinese showing "模型與服務費率" in traditional characters.
• Fixed German text switching between the informal du and the formal Sie, as in the chat list reading "auf Ihrem Mac" under "Was soll dein Mac tun?"; it says du throughout.
• Fixed French text switching between the formal vous and the informal tu, as in Settings › Dictation reading "ce que tu dis"; it says vous throughout.
• Fixed Portuguese on the Mac mixing in Brazilian words, as Sync, Scheduled Tasks, Teams and the design canvas saying "equipe", "arquivo", "salvar" and "excluir" next to "Definições"; they say "equipa", "ficheiro", "guardar" and "eliminar".
• Fixed Portuguese switching between the formal "o seu" and the informal "o teu", as on the xAI page offering "Usar o teu plano" next to "o seu plano"; it says o seu throughout.
• Fixed Portuguese addressing the reader as "você" and starting sentences with a bare "Seu" or "Sua", as the Agent Phone settings saying "Uma tarefa precisa de você"; it says "si" and "o seu" as European Portuguese does.
• Fixed Portuguese on iPhone and iPad mixing Brazilian and European Portuguese, as in Settings › AI Model saying both "Ajustes" and "Definições" or both "login" and "sessão"; it reads as European Portuguese throughout.
• Fixed the iPhone app's Portuguese permission prompts for the camera, location, contacts, calendars, reminders, Home and alarms using Brazilian wording such as "câmera" and "código de pareamento"; they now read as European Portuguese.
• Fixed Settings › AI Model's sign-in choices on iPhone and iPad in German and French, where In use left no room for the choice's name; In use now moves under a name it would cut.
• Fixed the New chat button at the bottom of the iPad chat list in French, where the pencil hung off the button and "Nouvelle discussion" was cut off.
• Fixed the iPhone app's Home showing "Ask anything" in English in the message box when the app uses another language.
• Fixed the iPhone app's chat list, Remote, open chats and sheets, and a chat's messages, such as file cards, changed-file summaries and the buttons under a reply, keeping some labels in the previous language after you change the app's language in Settings.
• Fixed VoiceOver reading many of the iPhone app's buttons, such as a chat's Side Panel, Chat options, Add photos and Dictate, and the Back, Forward, Reload, Share, Find and More buttons of a side panel page, in the previous language after you change the app's language in Settings.
• Fixed VoiceOver reading the iPhone app's sidebar button in English when the app uses another language.
• Fixed VoiceOver announcing in English, in every language, that a chat moved to or out of a project, that an invite code was copied, or that a file arrived during a call.
• Fixed the iPhone app's file changes card showing its file count in English, and VoiceOver reading its files and line counts in English, when the app uses another language.
• Fixed the iPhone app naming a web page, window, app or messages added to a message in English in every language.
• Fixed the iPhone app's settings showing their status, pairing errors, notices and dates in English in every other language, the pairing steps naming a button the iPhone no longer has, and copy confirmations for plugin routes, invite links and phone app links reading in English.
• Fixed some Worktree messages showing in English in every language, such as the warning that a terminal still has one of a Project's chats open when you archive or delete it, and the reasons a Worktree can't be archived or restored.
• Fixed the chat's tool summaries, such as "Read a file, ran commands", the "Working for", "Worked for" and "You stopped after" lines, the chat's message buttons, the Permissions menu and tool steps (search counts, approval buttons, output labels) staying in English when Enconvo is set to another language.
• Fixed chat error and status messages (attaching files, sending, editing, forking, compacting and inserting prompts) showing in English when Enconvo is set to another language.
• Fixed the message box's slash menu and the model chip's reasoning level (Low, Medium, High, and the rest) showing in English when Enconvo is set to another language.
• Fixed the chat's Install card and the app context chips above the message box showing English text when Enconvo is set to another language.
• Fixed the menus and messages of visualizations and cards in chat (copying, downloading, resetting values and the download dialog) showing in English when Enconvo is set to another language.
• Fixed agent question and approval cards showing their hints, placeholders, "asks" heading and answered labels in English in every other language.
• Fixed the conversation points pill's Context and Usage tabs showing English group names, counts and notes when Enconvo uses another language.
• Fixed Archived Chats (the settings page, the sidebar panel and its menu), the archive and restore messages, and the workflow canvas's input and end steps showing in English, and the archived chats' dates ignoring Enconvo's language.
• Fixed the Memory page (its filters, the box for adding a Memory, the details panel and who reads each Memory) showing in English, and its dates ignoring Enconvo's language.
• Fixed the saved post preview, an agent's Memory settings, the changes review, the Claude Code usage limits and the archived chat banner showing English in every other language.
• Fixed Ask AI in the knowledge base, and the chat menus of the workflow Assistant, Plugin Studio and scheduling chats, showing English lines and times when Enconvo uses another language.
• Fixed Ask AI in the knowledge base asking its suggested questions and Explain in English, which could bring the answer back in English, and showing citation places and errors in English when Enconvo uses another language.
• Fixed the lines under Ask AI answers that say what was read and searched, and its status while it works, showing in English when Enconvo uses another language.
• Fixed knowledge base messages (renaming, moving files to the Trash, pasting files, relinking a folder, saving transcripts), knowledge base items' type, date, size and status (card labels, the sort and Types menus, the list columns and the details panel), and a Project's Finish setting notes showing in English when Enconvo is set to another language.
• Fixed the MCP servers page's messages (connecting, installing, updating, importing and signing out of a server), its AI model access choices and its version notes showing in English when Enconvo is set to another language.
• Fixed the Skills page's messages (installing, uninstalling, importing and moving skills to the Trash), its switch names and the detail panel's source and date showing in English when Enconvo is set to another language.
• Fixed the Plugins page's subtitle, its Connector messages (connecting, disconnecting and changing the default account) and its buttons' names showing in English when Enconvo is set to another language.
• Fixed a Connector's page and its row on the Plugins page showing English or mistranslated, spliced sentences in other languages, such as the line about turning off macOS access after Disconnect.
• Fixed the card an agent shows to connect a Connector, its Gmail Mail Alerts question, and the API key form staying in English when Enconvo uses another language.
• Fixed the Computer Use settings showing English permission cards and browser connection times when Enconvo uses another language.
• Fixed the OpenClaw, Hermes and Ollama setup guides in AI model settings (their steps, statuses, setup dialogs and messages) showing in English when Enconvo is set to another language.
• Fixed model routes, tag management, reply details, the workflow overview, the knowledge source card and the skill install dialog showing English in every other language, the Command Return shortcut reading as "command result" in German, French, Spanish, Italian and Portuguese, and Chinese leaving the Dynamic Island untranslated on some screens.
• Fixed the X post saving settings, knowledge base filters and categories, the workflow step palette, the Smart Bar appearance setting and Live Talk's napping status showing English in every other language.
• Fixed IM Channels' connection messages, the knowledge base editor's word count, read-only and save messages, and New Knowledge Base and Screen Doodle settings messages showing in English when Enconvo is set to another language.
• Fixed Settings › Account's plan details, the Appearance, refund and log out messages, Settings search messages, the log out and remove Project dialogs, and the Cancel button of several confirmations showing in English when Enconvo is set to another language.
• Fixed the plans dialog and Settings › Permissions showing in English in other languages: plan cards, the benefits table, the buy, upgrade and add-seat buttons, and each macOS permission's name and purpose now follow the app's language, with point counts written the language's way.
• Fixed the Points page showing its history, daily usage chart and savings line partly in English or in a garbled word order in other languages, with its dates and numbers now written the way the app's language writes them, and Chinese calling points 点数 everywhere.
• Fixed Manage devices, Invite Friends, Redeem Code and About showing parts of their text in English or in English word order in other languages, writing dates and counts the language's way, and fixed Korean sentences that ended with a Chinese full stop.
• Fixed the icon picker, Raycast extension installs from GitHub, device limit notices, the redeem page and the MCP bundle installer showing English in every other language, and Portuguese addressing you informally on some screens, as "Tenta novamente" after an error.
• Fixed the Close button label of dialogs, the sidebar toggle, page buttons in Points usage and Heartbeat logs, workflow task states, and a few chat and speech model errors showing in English when Enconvo is set to another language.
• Fixed settings dropdowns showing a garbled empty-search line and English search hints, and settings text and number fields showing an English "Enter …" hint, when Enconvo is set to another language.
• Fixed page lists showing "Next step" instead of "Next page" in Chinese, the local model's memory usage showing as the Memory feature, German calling the Memory feature "Speicher", and page and count lines such as "Showing 1–20 of 80" keeping English word order in other languages.
• Fixed lines in chat, skills, MCP install, heartbeat, credentials, subscription usage, avatars and knowledge bases showing English word order or stray words in other languages, such as Chinese file counts ending in "秒", and wrote their times and dates in the app's language.
• Fixed searches with no match, delete and authorization prompts and other settings lines in languages other than English, which split a quoted name or a count from its sentence, such as Chinese reading 未找到匹配“的文件 before the search term.

Artifacts and Docs

• Added Artifacts: pages, small apps, diagrams and programs the agent makes open beside the chat with numbered versions you can restore, edit or download, and the Chat Window sidebar lists them all.
• Added Docs: ask for a report, plan or letter and the agent writes a doc beside the chat that you edit with tables, images and charts, comments, Suggesting and earlier versions.
• Added asking Enconvo inside a doc: type @ on an empty line, select text and choose Edit with Enconvo, or mention @Enconvo in a comment.

Projects and Worktrees

• Added members to Projects: give a Project a Lead and members, your agents and Coding agents each with a role, or start from a template, and new chats in the Project start with its Lead.
• Added Work plans: the Lead shows who does what, which you can change or reassign before Start, then runs each task in its own chat on the Board, and you can hand work to a Project from any chat or the @ menu.
• Added Playbooks: the built-in Build, Fix a bug, Review and Research Playbooks, or your own from the Workflow editor, take a task step by step, with the member you choose for each part.
• Added reviews, checks and recovery to Work plans: a task that changes files is reviewed by another member, and tasks cut short, stalled or going in circles run again or wait for you.
• Added a Needs you tray, Allow in this Project, a points budget, Pause and Stop plan to Project chats, with a Run log of each decision the Lead made.
• Added a Boundary to each Project that keeps its agents to the folders, Mac control, Connectors, MCP servers and knowledge bases you pick, even with Full access.
• Added Projects at work to the Dynamic Island, turning amber when a member needs you so you can answer in place.
• Added Worktrees to git Projects: a chat can work in its own worktree on a new branch, a branch or a pull request, with its own ports and Prepare command, set up in Project settings.
• Added Finish to worktree chats: commit with a drafted message, merge or open a pull request, and archive the worktree once it's merged, with Restore to bring it back.

Side Panel and Doodle

• Added Doodle to the side panel: draw on a PDF, picture, file, artifact or web page, or point at a paragraph or table and comment on it, typed or spoken, and the numbered marks go with your next message.
• Added spreadsheets to the side panel: Excel, Numbers and OpenDocument sheets open with their sheets and merged cells, and you can copy cells out.
• Changed the Changes tab into a review, as in Codex, one turn at a time with unified or side-by-side diffs, opened from Review on a turn's files changed card.

Chats and Chat Window

• Changed the Chat Window to the Smart Bar's translucent look, following the Liquid Glass setting on macOS 27, with an Appearance option in Settings › Chat Window.
• Added Open in Chat Window to the Smart Bar and the Dynamic Island, and Continue in Smart Bar (⇧⌘O) to the Chat Window, which move a chat with its running reply.
• Added Compare answers to the + menu: pick 2 to 6 agents, a saved set or Second Opinion, and the chat's agent lays out where their answers agree and differ.
• Added quick archiving: archive chats from the sidebar, several at once, all finished ones from the ⋯ menu, or automatically after 3, 7 or 30 days.
• Added /loop to agent chats, as in Claude Code, which runs a prompt again on an interval or whenever the agent decides.
• Changed Automations to Scheduled, where a task's result comes back to the chat you set it up in, with a notification on your phones, and a task can pause until a date.
• Changed switching chats to show the whole chat at once at its latest message, and the sidebar's status marks to quieter ones, as in Codex.
• Added the PopBar to text selected in the Chat Window, and ↑ and ↓ in an empty message box to step through the prompts sent in that chat.
• Changed messages sent while a reply runs to show at once and queued messages to show their pictures and files, and fixed Edit on a sent message doing nothing.
• Fixed attaching an .ico, BMP, TIFF or HEIC image breaking a chat, and keys typed in Enconvo's windows going to the Dynamic Island.

Coding Agents

• Added Codex's own commands to Codex chats, such as /status, /diff, /usage, /mcp, /plugins, /hooks and /fork, with $ mentions of apps and skills.
• Added Plan mode and the Approve for me permission mode to Codex chats, and Hooks, Skills, Plugins, Apps and Personality to Codex's settings.
• Added a permission mode picker to new Claude Code, Codex, Grok Build and Antigravity chats, and model and effort pickers to Grok Build and Antigravity chats.
• Added Claude Artifacts and claude.ai Connectors to Claude Code chats on a Claude subscription, and Claude in Chrome and Output Style to Claude Code's settings.
• Added Grok Build's questions and plan approval to Grok Build chats, and your Antigravity plan's Claude models to Antigravity chats.
• Changed renaming a Claude Code or Codex chat to rename its session too, and Continue in Terminal to keep the options from the agent's settings.
• Fixed Coding agent chats that won't start because of a setting, which now point to the setting to change.

Memory and Knowledge Base

• Changed the Knowledge Base to open in the Chat Window instead of its own window, with Ask, a table view and Library views, and in under a second with thousands of items.
• Changed web pages and long files you save to a knowledge base or Memory to be read by sections, so agents get more of what they say.
• Changed the Memory page, now in the Chat Window sidebar too, to show every Memory in full, grouped by Pinned and by day or project, with filters and an editable detail panel.
• Added Concepts to Memory, filing your Memories under the people, projects, places and topics they are about, and a Map view of your Memories and knowledge base items.
• Added a Memory section to each Agent's settings to choose which Memories it reads and whether it saves new ones.
• Removed the Memory card at the end of a chat: Memory now learns quietly, and Settings › Memory lists each change with Undo, going back 90 days for Memory Tidy.
• Fixed Memory saving relative dates instead of the date, passing illnesses and things you asked it to forget, writing in a language other than yours, and saving while Memory is off.

Live Talk, Voices and Dictation

• Changed Live Talk to hear you as you speak with Soniox on your own key, AssemblyAI, Azure, ElevenLabs Scribe v2 Realtime, Volcengine and xAI, and with xAI to speak over one connection for the whole call, so answers start sooner.
• Added Kokoro to Text to Speech and Live Talk on Macs with Apple silicon and macOS 15 or later: free, offline voices in English, Mandarin and Japanese.
• Fixed Phonon seeming to switch between a man's and a woman's voice mid-reply, and changed it to download from Enconvo's own server, much faster in many places.
• Changed SenseVoice on your Mac to write punctuation and numbers as digits, and fixed Dictation with Soniox stopping with "Context is too long".

Agent Phone

• Added Agent Phone: when an agent's question waits while you're away from your Mac, your iPhone rings after the time you choose, and you answer by voice.
• Added Twilio and Telnyx to Settings › Agent Phone, so agents can call your phone number when your iPhone can't ring, with Test call and quiet hours.
• Added calls agents make for you, such as booking a table, through Twilio or Telnyx, keeping to what the agent wrote down and reporting back in the chat.

Agents in Your Apps and Files

• Added a swipe action to Computer Use, made its scrolls take about half as long, and fixed dozens of cases with clicks, keys, checkboxes and sliders.
• Changed agents to read and edit text in older encodings such as GBK, Big5 and Shift_JIS, and to search inside PDF, Office, EPUB and iWork files.
• Fixed many cases in agents' file, calendar, reminder, Apple Notes, Obsidian and Contacts tools, such as names with curly quotes and all-day events.
• Fixed web pages crashing in Chrome and Edge with "Aw, Snap!" while Browser Use worked in them, with the Enconvo browser extension 0.1.16.
• Removed the Tools tab from Agent settings, since every Agent already uses its built-in tools, your MCP tools and everything you install.

Smart Bar, Plugins and Connectors

• Added Twilio, Telnyx and Bland Connectors for sending messages, placing calls and managing phone numbers and AI voice agents in your accounts.
• Added an Airmail plugin that writes an email from what you dictate or type, opens it in Airmail's composer, and searches, replies to and files your mail.
• Added Pollinations to Image Generation, with Sign in with Pollinations and a limited free tier that needs no key.
• Added a hotkey for each app in Settings › Applications that opens it, brings it to the front or hides it.
• Changed the Smart Bar's Pin window back to a toolbar button, and list hotkeys such as Clipboard History to open the list at once.
• Changed Screen Doodle text to grow and wrap as you type, and fixed Screen Doodle closing by itself when the Dock or the menu bar showed.
• Fixed weather, Emoji Picker, the calculator, File Search, Clipboard History, the compress commands and Mistral OCR in many cases.

Pet and Avatar

• Added an Ask box to the Pet: click it, or drop files, text or images on it, to ask its Agent, and the answer appears above the Pet.
• Added your own Avatar characters: make a plush in Settings › Avatar, describe one for your AI model to draft, or add one from the OpenAvatar gallery.
• Removed the Puppy, Bear and Cat characters; the iPhone and Android apps now start with Duet, as the Mac does.

Models and Sign-ins

• Added GPT-6.1 Terra, OpenAI's new mid-range model, with your own key, a ChatGPT sign-in and on Enconvo Cloud, as the OpenAI provider's default.
• Added Kimi K3.1 Pro and K3.1 Pro Highspeed, Moonshot's newest models with a 1M-token context, on Enconvo Cloud and with your own key.
• Added a subscription's 5-hour and weekly limits to the model picker, and clearer sign-in progress in Settings › AI Model.
• Changed agents on Claude Haiku 5.5 to think at Low effort by default, so they answer in your language and get dates right more often.
• Fixed LM Studio chats on a model loaded with a short context, Gemini Web Search not searching, and ChatGPT plan usage failing to load.

iPhone and Android

• Changed Settings › API Keys in the iPhone app to Settings › AI Model, as on the Mac, with 13 more providers, OpenAI and Anthropic Compatible ones, and models added by ID.
• Added sign-ins with your ChatGPT, Grok, Le Chat, GitHub Copilot or Antigravity plan, so Phone chats use your plan's models with no API key.
• Added iPhone Apps to Phone chats for events, reminders, contacts, places, Home accessories and alarms, and agents on your Mac can ask your iPhone too.
• Added the side panel to iPhone chats, with web pages and, for a Mac chat, its Files, Changes, Side chats, MCP Apps, Terminal and Doodle.
• Added Compare answers, chats with a Project's Lead and its Work plan, and a Live Activity for a Project at work to the iPhone app.
• Added Nemotron 3.5 Multilingual and SenseVoice Small for Dictation and Live Talk, and Kokoro voices on iOS 18 to iOS 26.3, all free and on the iPhone.
• Added iPhone Duo layouts that put chats, docs, sources and files side by side on the open inner display.
• Changed iPhone voice calls to a Siri-style look, with light around the screen's edge that pulses with the voices.
• Added dropping photos and files on a chat, and sharing files of any size to a Mac chat.
• Fixed phones losing the Mac after a short Enconvo relay outage until you relaunched Enconvo.

Usage, Plans and Privacy

• Added an Activity log to Settings › Usage Statistics with every AI request, tool call and scheduled task, kept only on your Mac, and fixed cached tokens being counted twice.
• Added the renewal terms below every Enconvo Cloud subscribe button, and changed the Affiliate program so Cloud points packs no longer earn a commission.
• Changed AI-made pages and shared chats to load fonts from Enconvo's servers instead of Google Fonts, and shared chat links to leave out steps' details and your folder paths.

Reliability

• Added Submit Diagnostics… to the menu bar menu, which sends recent logs and crash reports with personal details taken out, with each report in Settings › Diagnostics.
• Fixed Enconvo freezing for seconds when a slow app launched, when you selected text in a busy app, at launch, or while your microphone switched.
• Fixed Enconvo quitting when you started Dictation with AirPods, staying unresponsive after its engine got stuck, and crashing now and then when you sent a message.
• Fixed Enconvo keeping about 250 MB of memory after launch, loading its local embedding model after every launch, and Keep Awake staying on after you turned it off.

Memory

• Added Memory Review, which keeps your preferences, the people you mention and your plans as Memories once a chat or a Live Talk call goes quiet, and updates or forgets the ones it changed.
• Added Memory Tidy, which merges duplicates, settles plans whose date has passed and tidies tags once a day or once a week.
• Added Pinned Memories, which every agent always sees, such as an allergy or how you want to be addressed.
• Added recall across your Memories, past chats, recordings and knowledge bases in one search, keeping to the day, week or month you name.
• Added a Memory page in Settings to turn Memory on or off, undo what it learned or tidied, choose where agents recall from and pick its model, and a card in chat that shows what it learned, with Undo.
• Changed agents on a model running on your Mac to keep your past chats on your Mac, without sending passages of them, or Memory Review, to the cloud.

Live Talk, Voice Command and Quick Actions

• Added Live Talk and Voice Command pages in Settings, with a guide to each, to choose how it hears you, the model it talks with and its voice.
• Added Enconvo Fast, the default model for Live Talk and Voice Command: Claude Haiku 5.5 with thinking off, which answers right away, sees a photo, your camera or a shared screen, and is free up to $5 a day.
• Added Fish Audio as a Text to Speech provider and Live Talk voice, with Drama 3, S2.1 Pro and S2 Pro, free for now on Enconvo Cloud.
• Added Phonon, a free English voice from Gradium that runs on Macs with Apple silicon and macOS 15 or later, for Read Aloud, Live Talk and Voice Command, speaking almost at once.
• Added Settings › Quick Actions, which shows what Voice Command and Live Talk can run at once, and your own Quick Actions, made from a sentence or from built-in actions, Shortcuts, apps, links and scripts.
• Changed ElevenLabs to use Eleven v4 Turbo by default, and added Eleven v4 Turbo, v3 Conversational and Turbo v2 and v2.5 on Enconvo Cloud, with models and voices that refresh without an app update.
• Changed the ElevenLabs sound effects, music, dubbing and voice tools to need your own ElevenLabs API key; Enconvo Cloud still covers ElevenLabs speech and transcription.
• Changed Live Talk to answer sooner, with Smart Turn for every speech model, and added Parakeet and Fluid Speech Mini, which run free and offline on your Mac, to the recommended speech models.
• Changed tasks you hand off by voice to start sooner, with the app or web page in front already read.
• Fixed Voice Command missing your first words, and voice calls saying they did something without starting it, starting a task twice, or answering from an outdated state of your Mac.

Agents in Your Browser and Apps

• Added a glow, a cursor and a control center over the browser tab or app window the agent works in, with Take over, Stop and Return to agent.
• Added handoffs for steps only you should take, such as a CAPTCHA, a sign-in, a password or a payment: the agent waits, then goes on from where you left it.
• Added browser scripts and App scripts, so the agent goes through long lists, feeds and many pages in one step, and stops before anything that sends, deletes or buys unless you asked for it.
• Added reading the data a web page loads from its site, so the agent gets exact numbers and tables, and remembers how it reached a site's or an app's data for next time.
• Changed the agent to treat instructions in web pages, files and emails as content, never as commands, and to tell you when one asks for something you didn't.
• Changed Browser Use to handle dropdowns, date pickers, menus, rich text editors, uploads and pages that load as you scroll far more reliably, in tabs behind the one you're looking at too.
• Changed Computer Use to wait until you stop typing or clicking before it acts in a window, and to use what an app offers besides its window, such as AppleScript commands and URL schemes.

Workflows and Agents

• Added Judge and Classify steps to workflows, which answer yes or no, or pick one of your options, in about a second.
• Added workflows to Browser Use and Computer Use, with a repair agent that does a failed step another way so the workflow goes on.
• Added agent calls: any agent, Claude Code, Codex, Grok Build and Antigravity included, can ask another agent to do part of its work, in a Permission mode no wider than its own, and get the answer back.
• Added Connect AI Clients to the Agents page, so AI apps on your Mac, such as Claude Code, Codex CLI, Cursor and Claude Desktop, can ask your agents for work over MCP.
• Changed the agent to call its tools directly, so multi-step tasks finish noticeably faster.
• Fixed several ways an agent could reach past its Permission mode or skip asking before running a script or a command that deletes files.

Chats

• Added sliders, fields, choices, switches, progress bars, timelines and live data to the cards the agent draws in chat.
• Changed a finished agent reply to show only its answer, with its notes, thinking and steps folded into one "Worked for …" line, and each step reading as one line.
• Added a suggested next message in the empty message box after a reply; press Tab to use it.
• Added file links in agent answers that open in the side panel, and one gallery for a reply with several images.
• Added a model and reasoning effort picker to new Claude Code and Codex chats.
• Changed the Chat Window to show the agent at the top left, with a menu to switch it in a new chat.
• Added the pictures and files an Agent makes to the Dynamic Island, the voice reply and the call card as soon as they're done.
• Changed chats to start replying sooner, and Enconvo Cloud replies up to a second and a half sooner after a break.

Recordings and Transcription

• Changed a recording's page to show the transcript with a player that marks when each speaker talks, beside the summary, action items, notes and Ask AI.
• Added free automatic summaries of recordings with key points, decisions and action items you can tick off.
• Added Transcribe on this Mac for a recording whose transcription stopped, such as when your Enconvo Cloud points ran out.
• Added Live Translation to recordings with a model that can translate, such as Soniox Realtime.
• Changed the Recordings list to rows grouped by day with a Needs attention filter, and the start view to change the microphone, call audio and model in place.
• Fixed many issues with renaming recordings, naming speakers, deleting recordings and transcribing them again.

Smart Bar, Plugins and Mini Apps

• Changed the Smart Bar, the Small Smart Bar and the App Sidebar to the look of macOS Tahoe's Liquid Glass in Light and Dark Mode, with an Appearance option in Settings › Smart Bar.
• Changed list Mini Apps such as Clipboard, File Search and Bookmark Search to open at once, search as you type, page through every item, and filter with a menu beside the search field.
• Added a Built-in tab to Plugins with settings pages for Clipboard History, File Search and Browser Bookmarks, and Shortcut and Alias settings for each Mini App.
• Added an on/off switch to installed plugins, and switches for where each plugin's views show up.
• Fixed File Search taking seconds for each letter without Full Disk Access; results now show as you type.
• Changed Smart Bar search to rank what you use higher.

Hotkeys and AI Commands

• Added Other Hotkeys and Find by hotkey to Settings › Shortcut, which list every hotkey and what uses a key combination, with a clash warning and Use it here.
• Changed every hotkey field to one recorder that takes key combinations, modifier keys pressed together, and Hold, Tap, Tap, Hold or Double Tap.
• Changed Enconvo's default global shortcuts to only Smart Bar, Chat Window, Dictation, Voice Command and App Sidebar.
• Added All Commands to Settings › AI Commands: one model for every AI Command, writing tool, PopBar button and Dictation's Polish and Translate, set to Claude Haiku 5.5 on Enconvo Cloud.
• Changed AI Commands and writing tools to answer in about a second, keep your text's language and reply with the result alone.
• Fixed the PopBar not showing for text selected in Google Docs and other web editors that draw their pages on a canvas.

Dictation

• Added a choice of where Dictation pastes when you switch apps while dictating.
• Fixed custom Dictionary words with Parakeet, and pressing another key while holding the Dictation key cancelling the dictation.
• Fixed the first Dictation after Enconvo opens starting late, and stray symbols while a terminal app was in front.

Dynamic Island and Avatars

• Added a task list to the Dynamic Island: a Tasks chip opens a list grouped into Needs you, Failed, Running and Done.
• Changed the Dynamic Island settings to Home, Opening & Display and Character tabs, with Home shortcuts you can drag, rename and add.
• Changed the onboarding guide's host to Duet, Enconvo's own character, with an upbeat new voice.
• Added Live-Talk Voice to the Avatar's right-click menu, and the Avatar now acts on Voice Command replies and does the motions you ask for right away.

Models and Local Models

• Added Mercury 3 on Enconvo Cloud, Inception's diffusion model that writes over 1,000 tokens a second.
• Added Reasoning Effort for Mistral Large 4, Medium 3.5 and Small 4, and fixed Mistral's model list and Large 4 chats.
• Removed GPT-OSS 120B, GPT-OSS 20B and Gemini 2.5 Flash Lite from Enconvo Cloud; Mistral Small and MiniMax M3.1 Flash are the free models.
• Added a Local Models page in Settings that shows what the models on your Mac are doing, with Pause, Run everything in the cloud, and a choice of what reads text in images.
• Added Settings › Memory › Search, where the search index can be built free on Enconvo Cloud instead of your Mac's GPU, and the knowledge base index upgrade no longer keeps the GPU busy for hours.
• Changed local model downloads to ask whether to switch to a model once it finishes.
• Changed the Enconvo Cloud model list to check for new models and prices every hour.

Mac Apps

• Changed Calendar, Contacts, Mail, Messages, Music, Notes, Reminders and Maps into separate plugins on the Mac Apps tab, each turned on or off on its own.

iPhone and Android

• Added a Side chat and Live Talk settings to the iPhone app, with Parakeet, Fluid Speech Mini and Phonon hearing and speaking on the iPhone itself.
• Added Start a Voice Chat for the Action Button, the Shortcuts app and Control Center on iOS 18, without opening the app.
• Added a viewfinder and camera control to Share Camera in iPhone voice chats.
• Changed Mac chats on the iPhone to fold the agent's work behind "Worked for …" and offer a suggested next message, as on the Mac.
• Changed agent approvals and questions on the iPhone to arrive as Time Sensitive notifications, and on iPhone and Android to show where they were answered once you answer them elsewhere.

Reliability

• Added crash and freeze reports on the Mac and iPhone while sharing analytics is on, without any of your content.
• Fixed Enconvo freezing until you restarted it when its background engine stopped.
• Fixed Retry deleting a chat's earlier messages, Clear All deleting archived chats, and settings and API keys being erased when saved at the same moment.
• Changed the built-in Node.js runtime to Node.js 26.12.1, with the latest security fixes.

Automations

• Changed Scheduled to Automations, with filters, a Run box for tasks you start by hand, and switches for Important mail alerts.
• Added triggers for folders, apps, Apple Mail and Gmail, Discord and Telegram messages, MCP server events and transcribed recordings.
• Added web page and feed, clipboard, calendar, meeting, login and webhook triggers, and Mac triggers for waking, unlocking, plugging in and time away.
• Added conditions you describe in your own words, which AI checks for each event, even in a screenshot or scanned PDF, with Try on recent events.
• Added running automations by hand from the SmartBar, the PopBar and buttons on the Dynamic Island, which also shows each run's progress.
• Added an Approval step to workflows that waits up to 24 hours for your answer, and a step-by-step view of each run with Retry from the failed step.
• Added retries for steps that only read or ask a model, grouping of events that come close together, an hourly limit of 20 runs, and a loop guard.
• Added a Needs you card for failed events, pausing after 5 failed runs in a row, and a workflow of your choice to run when an automation fails.
• Added Make it run the same way every time, which turns a reliable agent automation into a workflow, and Undo for changes made in chat.

Coding Agents and Orchestrator

• Added Codex as a coding agent with your ChatGPT sign-in, its own models, reasoning effort and permission modes, and your plan's usage limits.
• Added subagents, file change cards with Undo, background commands, generated images, ! commands, /compact, /review, /goal and forking to Codex chats.
• Added Grok Build and Google Antigravity as coding agents, and Enconvo's skills and actions in Codex, Grok Build and Antigravity chats.
• Added one-click Install and Update for Claude Code, Codex and Grok Build without Terminal, with your account, plan and usage on their settings pages.
• Added a project's Codex, Grok Build and Antigravity sessions to the new-chat page, including ones started in a terminal, to continue as chats.
• Added Continue in Terminal or iTerm to the ⋯ menu, and Copy command for any other terminal, in Claude Code, Codex and Grok Build chats.
• Added the Orchestrator Board, which plans a Project's tasks and runs each coding task in its own worktree, to merge or open as a pull request.

Chat and Side Panel

• Added a brand-new side panel to the Chat Window, Smart Bar and Dynamic Island, with tabs for Files, web pages, a Terminal and Side chats.
• Added Side chats with /btw for a quick question while the chat keeps working, and a Changes tab with Undo for the files each turn changed.
• Added web pages in the side panel that the agent can open and use to check its work, asking you before paying, sending, deleting or publishing.
• Added diagrams to chat replies and Markdown previews, drawing Mermaid blocks in the chat's colors with a button to switch to the code.
• Added web page cards for HTML pages in replies, HTML previews that run their scripts, and editing a file's source in the side panel with ⌘S.
• Added a conversation map along the edge of long chats, a mic button in the Chat Window's message box, and Up to bring back what you sent before.
• Changed chat search to return results several times faster without missing older chats, and added output speed to each reply's Details.

Smart Bar, PopBar and Mini Apps

• Added Mini Apps: plugin apps open in the Smart Bar with a back button and an Actions menu (⌘K), and plugins can show searchable lists there.
• Added File Search and Browser Bookmarks Mini Apps, and Clipboard History now opens as a Mini App with ⌘1–⌘9 to paste.
• Added Auto-switch Input Source to the Smart Bar, and removed its Use Application as Context First and Shake to Dynamic Island options.
• Changed the Smart Bar to list your recent commands and apps near the top, find apps by names in other languages, and offer Quit App and Force Quit.
• Fixed the PopBar hotkey to hide the PopBar when pressed again and show it under the selected text, and Pinyin input in the PopBar's Ask field.

Computer Use

• Changed agent mode to act in a Mac app's window right away and see only what changed after each step, so desktop tasks start sooner and use fewer tokens.
• Added right-clicks to the Computer and Browser Operators, and a screenshot fallback for controls and lists an app shows only on screen.
• Fixed Computer Use in Spotify and other web-based apps, so menus, drop-downs, sliders, checkboxes and long lists work as asked.
• Fixed Computer Use typing into background apps such as the App Store and into spreadsheet cells, and reporting actions as done when nothing happened.
• Removed the Computer Use and Browser Use MCP servers for other AI apps such as Claude Code, Codex and Cursor, along with Connect AI Clients (MCP).

Dynamic Island, Read Aloud and Avatars

• Added activity chips to the Dynamic Island's Home for downloads, Read Aloud, working Agents and waiting questions, and two activities on the notch.
• Changed Read Aloud to open its controls when you click the notch, hold its place during a Voice Command or call, and leave the Avatar uncovered.
• Added Natural lip sync on Apple silicon Macs and the iPhone, an on-device model that moves an Avatar's lips and tongue with the voice she speaks.

Voice and Dictation

• Changed Voice Command on the Mac to open apps, run quick actions and start browser searches as soon as you finish speaking, two or three at a time.
• Added browser search to Mac voice calls, and calls now find apps by names in other languages or the names people call them.
• Fixed Voice Command talking over you when you speak again, and voice calls pausing when Enconvo Cloud's ElevenLabs voice runs out of credits.
• Changed On-Device Cleanup to Enconvo Tidy 1, which also cleans up Chinese, Japanese, Korean, German, French, Spanish, Portuguese and Italian.
• Changed On-Device Cleanup to work on macOS 14 too, and removed its Tone setting and the Spokenly Cleanup 1 download, freeing about 730 MB.
• Fixed Dictation in Pages moving the cursor, crashes when Dictation read Enconvo's own window, and permission alerts hiding behind other windows.

Recordings and Downloads

• Changed meeting recording to happen on Settings › Recordings, with a live transcript, Ask AI during the meeting, notes, bookmarks and Pause.
• Added an Audio sources window for the microphone, call audio and transcription model, a no-sound warning, and WeChat calls in the Meeting reminder.
• Added video and audio download progress to the Dynamic Island, with Cancel, a percent ring on the notch and Show in Finder when done.
• Changed video downloads to start in half the time, fetch long videos up to three times faster, save to chat history and reuse a file you already have.
• Fixed YouTube links from a playlist or Mix downloading the whole list.

Local Models and Knowledge Base

• Changed the MLX provider into Local LLM, a powerful, fast built-in engine for macOS 15 and later, with live memory and speed charts and a best model picked for your Mac.
• Added Hugging Face model search, Bonsai 2 models with vision, Thinking and Unload when idle, and Qwen3.5 4B as the default local model.
• Changed local model downloads to resume after a restart, retry on network drops and check free disk space before they start.
• Changed the local Anthropic and Responses endpoints to read PDFs, follow a JSON schema, and answer Claude Code on a local model much faster.
• Changed chat to search a long knowledge base item or recording you @-mention, so local models answer in seconds, and follow-ups keep searching it.
• Fixed agents on local models losing their tools, and knowledge base answers switching to English or guessing when a search timed out.

Models and Enconvo Cloud

• Added Antigravity and GitHub Copilot providers with Google or GitHub sign-in.
• Added Mistral account sign-in for Mistral AI, Voxtral and Mistral OCR.
• Added OpenRouter as a text-to-speech and speech-to-text provider, so its speech models work with one OpenRouter API key.
• Added MiniMax M3.1 Flash on Enconvo Cloud and with your own key, at half the price of MiniMax M3 and about three times faster.
• Changed provider settings to show API Key, Subscription and Enconvo Cloud as sign-in cards, with a Cloud points button on Enconvo Cloud providers.
• Added DeepSeek to Web Search and Web Fetch; past the free monthly quota, Enconvo Cloud search uses it at 35 points per search with results.
• Added Enconvo Cloud to the agent's ElevenLabs tools, such as sound effects, music and dubbing, so they work without your own ElevenLabs key.
• Removed the Temperature setting from every AI model provider, and Gemini AI OAuth and Qwen account sign-in, which now take an API key.
• Removed the Japanese voices from MLX Kokoro, which always failed to speak.

Plugins and Connectors

• Added Plugin Studio, where Create plugin opens a chat that builds plugins and PopBar extensions with you, ready to use or publish to the plugin store.
• Added automatic updates for plugins from the store, GitHub, npm, marketplaces and the PopBar catalog, holding back ones that change hooks you allowed.
• Added Comments and Report to plugin store pages, and review status, Publish again and Withdraw to the plugins you published.
• Added entry points for plugins with their own interface in the Chat Window sidebar, the side panel, SmartBar search, @ and the PopBar.
• Added a Location & Maps plugin for travel times, when to leave and places nearby or anywhere, using Apple Maps with no API key.
• Changed uninstalling a plugin to also delete its data, unless you add --keep-data to enconvo plugin uninstall.
• Removed installing .dxt files; MCP bundles now install only as .mcpb files.
• Added about 300 new Connectors, including Greenhouse, GitBook, DigitalOcean, Runway, Luma, Kling AI and Canny.

Sync and Settings

• Added iCloud as a Sync storage option for paid members, with changes from your other Macs often arriving within seconds.
• Changed API key sync so a new computer is approved from one that already has your keys with a six-digit code, instead of a passphrase.
• Fixed Sync losing edits made while it was off, bringing back deleted items, and replacing newer settings or API keys with older copies.
• Changed Settings search to open results without a pause, keep typing smooth, and rank pages and whole-word matches first.
• Fixed web pages shown inside Enconvo being able to send commands to the app, and plugin names or synced settings changing files outside Enconvo's folders.

iPhone and Android

• Added Codex chats to the iPhone app, and a project's Codex sessions to the iPhone and Android apps, including ones started in a terminal.
• Added message search to Search chats in the iPhone app, finding what was said in your iPhone's and Mac's chats, with previews.
• Added hiding the Avatar during voice chats on iPhone and Android, and web page cards that open HTML pages full screen in the iPhone app.
• Changed Mac chats in the iPhone app to keep typing and sending through a brief connection drop, and to show your first message the moment you tap Send.
• Fixed text in the iPhone app showing in English in other languages, across chat lists, search, settings, pairing, errors and VoiceOver labels.

Agents and Chat

• Added an Auto run mode for Agents, now the default, which answers quick questions directly and runs tasks with tools, replacing the Run Mode switch.
• Changed Agents settings to a card gallery where agents can be paused and resumed, and removed the Tools picker and Tutor Mode.
• Changed Steer to answer your new message right away, and added Cmd+Enter to steer every queued message into the running task.
• Changed agents to find your files by name in about a second, open the pages you name straight away, and ask before changing settings you didn't mention.
• Added recommended answers to agent questions, chosen after three minutes.
• Added Rename, Pin, Fork and Archive to the Chat Window's ⋯ menu, a Details button under each reply, and Copy as Rich Text, Markdown or Plain Text.
• Added Response Language to Settings › General, so agents, voice calls and the Avatar always reply in the language you choose.
• Changed My Profile to About You, now under Agents in Settings.
• Changed shared chat links to look the same as the chat in Enconvo, with charts, citations, media and tool steps that open as in the app.
• Fixed agents switching to English after using tools, replies showing words twice, and attachments with unusual file names.
• Fixed the chat history looking empty after an update until Enconvo restarted.

Voice Calls and Voice Command

• Changed voice calls to run on Enconvo Voice, which hears you with Soniox and answers in an expressive ElevenLabs voice with live captions.
• Added 42 ElevenLabs voices in 12 languages for calls, Read Aloud and Avatars, and each Agent now speaks in its Avatar's voice.
• Changed voice calls to search the web, read files, control your Mac and handle music, reminders and notes on their own, several at once.
• Changed Mac voice calls to work like the iPhone's, with words that appear as you speak and Screen Doodle and Live Screen sent to the call.
• Removed Auto Hang-Up, so a quiet call naps until you speak instead of ending.
• Changed voice calls to remember much longer conversations and recognize the words in your Dictation dictionary.
• Added spoken answers to push-to-talk Voice Command, which knows which app you're using when you speak.
• Changed Voice Command and voice call replies to be shorter and arrive sooner.
• Changed voice calls to try what you ask before saying it can't be done, doing it in the browser or handing it to the Agent.
• Fixed voice calls claiming a task was done without doing it, missing your corrections, giving tasks the wrong date, and ending when speech recognition dropped.

Dictation and Recording

• Added Edit selected text by voice, so you can select text and say how to change it, such as "make this more formal".
• Added Paste Last Dictation, Heard as entries and suggestions in the Dictionary, and an Overview in Dictation History.
• Added App Rules for the post action, language and Return key in each app or website, plus a Dictation Language setting.
• Added Fluid Speech Mini as an on-device English transcription model on macOS 15 and later, for dictation and live captions.
• Added On-Device Cleanup, a post action that removes filler words and fixes punctuation offline on Apple silicon Macs with macOS 15 and later.
• Changed Settings › Dictation to follow the steps of a dictation, with Clean up options whose prompts and models you can edit, or add your own.
• Changed Dictation to show your words sooner, stay ready for five minutes, and keep the text on the clipboard when no field is focused.
• Changed Silence other audio while recording to pause music and videos and resume them when you finish.
• Added recording from second-channel microphones, and meeting recordings now continue when the microphone changes.

Updates and Notifications

• Added Updates, which holds agent questions, approvals and finished tasks while you're busy and sends them to your iPhone once you leave the Mac.
• Added Pause Updates for 1 Hour and Quiet Hours to the menu bar and Settings › Notifications.
• Added Approve and Deny buttons to approval notifications, and changed the Dynamic Island to wait for you instead of opening for a question.

SmartBar, PopBar and AI Commands

• Changed the SmartBar to show results sooner, redraw far less and follow the app language throughout.
• Added PopBar extensions in PopClip's format from Enconvo's reviewed catalog, a link, a file or your PopClip.
• Changed the PopBar with a searchable More panel, an Ask panel for any AI Command, info buttons, button groups and a result card.
• Added Save to Knowledge Base as a default PopBar button, with its progress on the Dynamic Island and links saved as the page itself.
• Added New AI Command from a description, Customize for the built-in commands, and Try it on your own text.
• Changed Translator to join the Writing Toolkit and follow your Response Language, with Source and Target pickers.
• Removed the DeepL, Youdao, Caiyun and Translator (Interactive) commands, the AI Command template store, and AI Edit on the PopBar.
• Added spoken labels to Screen Doodle for the shapes you draw.
• Changed Screenshot OCR to open its result on the Dynamic Island.

Dynamic Island, Read Aloud and Avatars

• Changed Read Aloud to play on the Dynamic Island one sentence at a time, with captions and speed and voice pickers.
• Added reading aloud by the Avatar, in her own voice with lip sync.
• Added eighteen built-in Avatar characters on the Mac and iPhone, including six Enconvo plush originals led by Duet, the new default.
• Added dances, motions, outfits and working moods to the Avatar characters, plus head pats and pointer tracking on the Mac.
• Added Island Character, which shows your Avatar in the compact Dynamic Island, and an All Screens option, now the default.

Music and Mac Apps

• Added Spotify and Apple Music plugins, so agents and voice calls can find, play and organize your music without signing in.
• Added a Mac Apps tab to Plugins, and Spotify and Apple Music buttons on the Dynamic Island that listen for what to play.
• Changed Dance Along with Music to be on by default.
• Fixed Play/Pause and track controls doing nothing, and Spotify taking up to 25 seconds to play a song by name.

Browser Extension and Knowledge Base

• Added Ask Enconvo to the browser extension on X, Reddit, YouTube, Gmail, Google Docs, GitHub, Wikipedia and many more sites.
• Added Firefox and Safari support, and video and audio downloads from YouTube, X, Bilibili, TikTok, TED and Vimeo.
• Added an Enconvo button to X's composer for polishing, shortening and replying, and AI reply suggestions on Reddit and Facebook.
• Changed saving posts on X to use X's own bookmark button, with automatic categories for posts saved from any site.
• Added Tags to the knowledge base, with Suggest categories, Manage Tags, an Untagged filter and a filter bar.

Plugins, Connectors and Computer Use

• Added about 300 new Connectors, including AWS, Microsoft 365, Google Maps, Tableau, Zoho CRM and Sonos.
• Added installing Codex and Claude Code plugins from GitHub, npm, a folder or a marketplace, with their commands, agents and hooks.
• Added enconvo plugin commands to validate, pack and publish plugins.
• Added a Cloudflare CLI plugin and the EmbeddingGemma 2 local model.
• Changed install links to show what they would install before you confirm.
• Changed Computer Use to ask before pressing Delete, Send or Buy, hide password fields, and report actions that didn't take effect.
• Fixed a crash on macOS 27 when Computer Use read Enconvo's own windows.
• Removed the Straico plugin and its providers.

Claude Code

• Added Claude Code to the Chat Window, running the Claude Code installed on your Mac in any folder, with your Claude subscription or an API key.
• Added Plan mode, bash mode, background tasks, Fast mode, a context meter, slash commands and clickable file references to Claude Code chats.
• Added your Claude Code sessions to Enconvo, including ones started in Terminal, with Fork and Continue in Terminal or iTerm.
• Added Claude Code chats to the iPhone app, with the same steps and plan card.

Models and Usage

• Added Auto Route, the new default model, which runs each message on the model you set for that kind of work.
• Added Claude Haiku 5.5, now the default for Enconvo Cloud and Anthropic.
• Added Gemini 4 Argon, half price on Enconvo Cloud through November 30, 2026.
• Changed the model picker to two columns, with a Fast mode chip for supported models that runs them faster at a higher price.
• Changed Manage Models to add models in a form, restore deleted models, and load OpenRouter models faster.
• Changed the quick decisions behind Auto Route, Auto run mode and Memory to use no Cloud points.
• Changed Live Talk voice calls and Soniox dictation to be free for Cloud Plan, Lifetime and trial members.
• Added HeyGen Voice on Enconvo Cloud for Chinese and English speech.

iPhone and Android

• Added a Dictation page to iPhone Settings with Soniox and two offline models, Fluid Speech Mini and Parakeet TDT V3.
• Added Fast mode to the iPhone model sheet, and Read Aloud now uses the ElevenLabs voice picked on the Mac.
• Fixed notifications sent while the iPhone was disconnected being lost.
• Added Continue with Google back to the iPhone and Android apps, and account deletion to Android.

Account and Teams

• Added a one-time 7-day Plus trial for Free accounts with no card needed, and trial codes for 30 days of Plus Cloud.
• Changed Teams so each Seat goes to a teammate's own account, managed by the admin from a new Teams page.

Voice and Avatars

• Added desktop Avatars with their own appearance, voice, expressions, and gestures, including Puppy, Bear, and Cat.
• Added voice calls from Chat Window, SmartBar, and Dynamic Island, with live transcripts, screen and camera sharing, and optional automatic hang-up.
• Added spoken Voice Command replies with captions that follow the speech.
• Changed Dynamic Island to keep your words, the Agent's progress, and its answer visible while a voice task runs.
• Added Avatar appearance and motion controls, saved looks, and a lighter rendering mode that reduces resource use.
• Fixed voice calls losing their chat context, ending while waiting for your answer, or failing to pass along photos and files.
• Fixed Dictation startup, post-actions, recording recovery, and local speech model loading.
• Added richer Screen Doodle controls with more reliable screenshot and annotation handoff.

Sync and Memory

• Added Sync across Macs for chats and their attachments, Projects, knowledge bases, settings, AI commands, Agents, workflows, skills, and snippets.
• Added optional API key sync, encrypted with a passphrase and off by default.
• Changed Sync to recover more quickly after a Mac reconnects and to explain files or settings that stay on one Mac.
• Changed Sync to keep new chats flowing when storage fills by removing the oldest cloud copies while preserving copies already on your Macs.
• Changed Memory to recall relevant preferences and older facts more reliably, update information that has changed, and handle dates more accurately.
• Changed Memory settings to use the same card and list views as Knowledge Base.
• Fixed Quick Memory and voice-note saving, transcription, and cancellation.

Knowledge Base and Saved Posts

• Added Ask AI alongside documents, books, recordings, and saved posts, with citations that open the relevant passage or playback position.
• Changed knowledge base search to find exact names and codes as well as related ideas, with clearer source references.
• Added saving X posts and threads, including images, into your knowledge base.
• Added readable saved-post views, video transcripts, and questions about individual posts or collections.
• Added a Knowledge Base MCP server for searching your saved material from Claude Code, Codex, Cursor, and other compatible apps.
• Fixed importing, previewing, and searching documents and recordings.

Agents and Everyday Automation

• Added parallel subagents for independent tasks, with support for project helpers and visible progress.
• Added per-Agent skill selection, shared and workspace skills, and safer folder imports and deletion.
• Added permission choices, task checklists, file-change undo, message editing, and conversation forks.
• Added faster Browser Operator and Computer Operator modes that pause before sensitive actions.
• Fixed a conversation with a custom Agent unexpectedly switching to the default Agent while a task was running.
• Fixed typing, scrolling, file picking, browser tab control, and interrupted tool requests across Computer Use and Browser Use.
• Changed tools on Mac to use Enconvo's bundled Node.js for a more dependable environment, even when another Node.js is installed.

Connectors and MCP

• Added Connectors for Gmail, Google services, Outlook, Notion, Slack, and more everyday work tools, with guided account setup.
• Changed Agents to recognize your connected accounts so requests such as "email me" or "add it to my calendar" use the right account.
• Added local-file attachments to emails sent or drafted with Gmail and Outlook, with download links for files too large for an email.
• Added interactive MCP Apps inside chat and clearer server and tool controls.
• Fixed Gmail migration, account sign-in, credential refresh, and reconnection.

Workflows and Scheduled Tasks

• Added an Assistant that creates and edits workflows from a description, with templates and workflow YAML import.
• Added a searchable step palette, recently used tools, clearer step cards, run summaries, and individual step results.
• Added input forms, value pickers, and warnings for missing or unavailable inputs, with an Assistant to help fix failed steps.
• Changed the Workflows page to a list and details view with recent runs, scheduling, and quick actions.
• Added setting up scheduled tasks by chatting, with plain-language schedules, run history, notifications, and delivery to connected chats.
• Added scheduling saved workflows with their own inputs and links back to each run.

Chat and Sharing

• Added public or password-protected chat links and Continue in Enconvo to copy a shared conversation into the app.
• Added Share and Copy menus to Chat Window, SmartBar, and Dynamic Island, including Copy as Markdown and chat links.
• Changed new shared chat links to open at app.enconvo.com in the reader's language while keeping earlier links working.
• Fixed chat scrolling over code blocks, tables, and wide content, including on shared chat pages.
• Changed long conversations to stream more smoothly, with clearer code highlighting, message queues, and history navigation.
• Fixed pinning chats, restoring empty chats, and sidebar chat names.

Models and Usage

• Added support during this release for GPT-6.1 Sol, Claude Opus 5.5 and Sonnet 5.5, GLM-5.3 models, and other supported provider updates.
• Added an MLX Models page for downloading, loading, and choosing local models, with OpenAI- and Anthropic-compatible endpoints for other apps.
• Added prompt caching for local MLX models and more reliable cancellation.
• Added Response Timeout settings to LM Studio and Ollama, with no time limit by default for local models.
• Added Usage Statistics across AI providers with estimated costs, token and cache breakdowns, model details, and a request log.
• Added clearer credential labels and subscription usage in the model picker.
• Added Cloud points visibility per conversation and spending checkpoints during long Agent runs.
• Added more image, video, and speech providers, including ElevenLabs voices through Enconvo Cloud.

Setup and Companion

• Added a shorter narrated first-launch guide, settings search, and expanded interface languages.
• Added pairing with the iPhone Companion to continue Mac chats, answer Agent questions, and receive task notifications.
• Changed the iPhone Companion experience with native chats, voice calls, Avatars, and settings shared with your Mac.
• Changed iPhone Companion notifications about activity on your Mac to play the alert sound.
• Changed the free plan to monthly allowances for Recording, Live Captioning, and Live Talk, with clearer limits for scheduled tasks and connected devices.
• Changed Sync to require an Enconvo Cloud plan or lifetime license.
• Changed the Mac app to use Enconvo's company signing identity; when updating from 2.5.5, macOS will ask you to grant permissions such as Accessibility, Screen Recording, and Microphone again.

Dynamic Island And Everyday Interaction

• Added a live Voice Command HUD with a waveform, transcript, and clear recording and processing states.
• Improved hotkey response so recording starts immediately and captures your first words while the panel opens.
• Added screenshots, file attachments, Screen Doodle, and automatic read-aloud controls to Dynamic Island.
• Added Default Agent switching from the agent icon in SmartBar, Chat Window, and the Dynamic Island panel.
• Fixed commands returning their results to the wrong window.
• Improved App Sidebar opening, window following, closing, and remembered width for each app.
• Added Inline Mode for applying commands directly to selected text, with a compact progress indicator and the result pasted back in place.
• Improved SmartBar input, Chinese text composition, command ordering, and PopBar behavior around window dragging and design apps.

Voice, Recording, And Live Captioning

• Added CoreML Parakeet, SenseVoice, and Paraformer dictation models with download, load, unload, and delete controls.
• Added speaker diarization and recognition to meeting recording, with word-level transcript alignment during playback.
• Added Live Captioning with realtime transcripts, speaker labels, translation, local model support, and saved history.
• Fixed Voice Command on macOS 27, Dictation post-actions, and local speech model loading.
• Improved recording sounds, volume restoration, dictation translation, and the way spoken input merges with your typed edits.
• Improved Read Aloud reliability and refreshed xAI voice selection.

Workflows, Agents, And Visual Results

• Added a visual workflow builder with step configuration, live execution status, individual step runs, and YAML editing.
• Added AI-assisted workflow creation and editing using Enconvo capabilities, MCP tools, and supported VS Code tools.
• Added the Visualize plugin for SVG, HTML previews, and interactive widgets inside chat, alongside native tables, charts, weather cards, and other generated results.
• Improved workflow cancellation and output so runs are easier to follow and stop.
• Added immediate Steer messages so you can correct an active Agent task while its tools continue running, including messages from Telegram and Discord.
• Improved long-conversation memory, model switching, prompt caching, and cancellation of stopped requests.
• Added settings to open Dynamic Island or send a system notification when a task finishes, with fewer duplicate prompts for conversations already visible.

Projects, Knowledge, And App Context

• Added Projects with shared instructions, folder-backed chats, a file browser, and Git branch controls.
• Added file and folder mentions from your current project, alongside workflow, MCP server, Finder file, Notes, and Mail context.
• Redesigned Knowledge Base with a folder tree, previews, tags, an inspector, drag-and-drop imports, and a standalone window.
• Added Quick Memory: double-tap right Shift to save the current app context, or hold right Shift to record a voice note for your knowledge base.
• Improved imports and retrieval for PDFs, images, and larger knowledge bases.
• Added app presets and site Command Packs for more direct automation, plus improved Office, Finder, Browser Use, and Computer Use support.
• Added inline attachment previews and live progress for video and audio downloads.

Models And Cloud Plans

• Expanded model choice with GPT-6 Astra, the GPT-5.6 family, Grok 4.6, Gemini 3.8 Flash, GLM-5.3-Flash, Kimi K3, MiniMax M3, and DeepSeek V4.1 Flash through their supported providers.
• Added GPT Image 2.5 Flare and Sunburst for image creation and editing, with support for API keys, Enconvo Cloud, and supported ChatGPT sign-in workflows.
• Added Google Gemini and Kie.ai video-generation options and expanded xAI voice, image, video, and search support.
• Improved custom provider model lists, fuzzy model search, and saved choices.
• Added clearer Cloud plan comparisons and upgrades, with DeepSeek usage at half the standard point rate on Pro and one quarter on Max.
• Redesigned Points Usage with clearer balances, usage breakdowns, and monthly allowance reset history.

Settings And Reliability

• Added new Skills, Plugins, MCP Servers, and Permissions management pages.
• Improved MCP setup, OAuth sign-in, credential refresh, and reconnection.
• Added account device management and clearer recovery when you reach your device limit or move to a new Mac.
• Added the Pet desktop companion with live task cards and community packs.
• Added a help menu for companion setup, release notes, and keyboard shortcuts.
• Fixed high background CPU usage, slow app switching, Dock icon hiding, and workspace preview close controls.
• Improved chat syntax highlighting, scrolling, panel transitions, and startup performance throughout the app.

Dynamic Island

• Added an Enable or Disable Dynamic Island item to the menu bar menu.
• Improved Dynamic Island event handling to match SmartBar behavior more closely, including Escape, Cmd+W, Cmd+N, and Return key flows.
• Replaced raw event monitors with stack-based hotkey monitors for more reliable open and close cycles.
• Fixed repeated compact hover entries sometimes failing to open.
• Hover validation now respects the padded hover target, so activation is less fragile around the visible pill.
• Added an optional Exclude Notch Area setting. When enabled, hover inside the physical notch area will not trigger Dynamic Island, while hover below or beside the notch still works.
• Removed streaming text preview from compact and expanded Dynamic Island modes when agent auto-read is active.

Chat And Sidebar

• Added the Ask Sidebar command for sending selected text or file context into the app sidebar.
• Registered Ask Sidebar as a Finder service and general right-click action.
• Chat With AI now avoids fenced code blocks unless you ask for code, commands, structured data, logs, or exact copyable text.
• Formula-style answers now stay in normal prose or Markdown math instead of code-style boxes.
• Improved long chat rendering by stabilizing Streamdown props and using content visibility for off-screen cached messages.

Privacy And App Package

• Signed-in users are identified more consistently across the Mac app, Node.js runtime, and web settings app.
• Sensitive values such as API keys, passwords, and tokens are filtered before analytics events are sent.
• Event names are clearer for command runs, panels, providers, settings, login/logout, and feature flags.
• Removed unused Swift packages including OpenAISwift, ActivityIndicatorView, SwiftUIPager, OpenAI by MacPaw, Blackbird, GoogleSignIn, GoogleSignInSwift, MenuBarExtraAccess, MSDisplayLink, SpringInterpolation, and Splash.
• Deleted unused WindowAnimation helpers and cleaned dead Swift imports in views and services.

Models And Providers

• OpenRouter Claude Opus and Sonnet model ids now match reasoning effort profiles by model family and version, even when naming order or separators vary.
• Opus 4.7 and 4.8 now show adaptive Low, Medium, High, X-High, and Max effort choices instead of legacy token-budget choices.
• Other Claude fallback models, including Sonnet 4.5 and latest aliases, now use the simpler Low, Medium, High, and Max effort list.

Dynamic Island And Voice

• Added voice command input for Dynamic Island so you can capture an idea or ask for help without switching into the full chat window.
• Improved Dynamic Island session reuse, hover behavior, model picker behavior, handoff state, and app-context screenshots.
• Added voice input to Chat Window for a faster way to start or continue a conversation.
• Improved how conversations move between native surfaces and web chat so the same session stays consistent.

Chat Window And SmartBar

• Improved Chat Window history syncing, titlebar navigation, row actions, prompt mentions, paste routing, footer controls, and model picker spacing.
• Improved SmartBar command mentions, slash model picking, status handling, paste behavior, mention chip removal, and follow-up flow.
• Made long conversations and repeated quick actions feel steadier across both main chat surfaces.

Computer Use And Browser Use

• Improved native app control with faster app state loading, stronger app and window targeting, better action waiting, and safer background behavior.
• Improved clicking, coordinate clicking, dragging, scrolling, low-level mouse control, and native panel actions.
• Added better support for native file picker selection, app menu element indexes, clipboard access, and permission prompts.
• Improved browser automation with exact tab targeting, cross-origin iframe handling, cleaner page state output, drag-safe clicks, and clipboard proxy support.

Knowledge Base, Context, And Search

• Improved Knowledge Base navigation, access, local API behavior, and settings history navigation.
• Added a Knowledge Base context manager and clearer search context actions.
• Added Docling document loader support.
• Improved application and website mentions, local SmartBar mention icons, installed plugin mentions, and window context capture.

Models, Providers, And Media

• Added Claude Opus 4.8 Cloud Plan support.
• Added GLM-5.2 and MiniMax M3 Cloud Plan support.
• Added xAI Imagine image generation and xAI video local media upload support.
• Added xAI OAuth credentials, xAI speech-to-text, and improved Grok model visibility.
• Added OpenAI OAuth GPT Image 2 and OpenAI OAuth Transcribe support.
• Added Soniox v5 speech-to-text models.
• Improved Anthropic OAuth account info, account switching, cancel behavior, high reasoning defaults, and image fallback behavior.

Settings, Skills, And Polish

• Improved hooks configuration, manual documentation, sound notifications, and agent monitor callbacks.
• Improved Skills Manager multi-source install reliability.
• Simplified Settings sidebar labels for General and Shortcuts.
• Improved Settings layout, Knowledge Base settings navigation, provider rows, and reset behavior.
• Improved PopBar row density, keyboard dismissal, copy shortcut conflicts, App Sidebar empty states, and window exclusions.
• Fixed TTS provider voice preview selection and IM channel launch lock recovery.
• Improved tool-result compaction, multimodal tool results, agent prompt cleanup, thinking message behavior, and chat stream rendering.

Chat Window And SmartBar

• Added voice input to the Chat Window prompt, so you can dictate directly where you already type.
• Added an Update button in both Chat Window and SmartBar when Enconvo detects a new version.
• Improved Chat Window history syncing, titlebar navigation, row actions, prompt mentions, paste routing, footer controls, and model picker spacing.
• Refined message readability, text selection, realtime assistant history, and table rendering in chat messages.
• Improved SmartBar status handling, command mention routing, slash model picking, paste behavior, mention chip removal, and follow-up flow.

Update Experience

• Simplified update channels to Production and Beta.
• Beta releases now publish to the matching Beta distribution path and appcast, so users on the Beta channel can detect the correct update.
• Update checks now run when Chat Window or SmartBar is shown, with a 10-minute throttle to avoid repeated background checks.
• Automatic update downloads are off by default.

Voice, Dictation, And Dynamic Island

• Added Dynamic Island voice command input.
• Improved voice command session reuse, so follow-up voice commands feel more continuous.
• Added xAI speech-to-text provider support.
• Added OpenAI OAuth Transcribe support.
• Added Soniox v5 speech-to-text models.

Computer Use And Browser Use

• Improved native app control with faster app state loading, stronger app and window targeting, better action waiting, and safer background behavior.
• Improved clicking, coordinate clicking, dragging, scrolling, low-level mouse control, and native panel actions.
• Added better support for native file picker selection, app menu element indexes, clipboard access, and permission authorization prompts.
• Improved browser automation with exact tab targeting, cross-origin iframe handling, cleaner page state output, better readable snapshots, drag-safe clicks, and clipboard proxy support.

Knowledge Base, Context, And Search

• Improved Knowledge Base navigation, access, local API behavior, and settings history navigation.
• Added a Knowledge Base context manager and clearer search context actions.
• Added Docling document loader support.
• Improved hierarchical local API schema lookup and API schema explorer runtime paths.

Models, Providers, And Media Generation

• Added Claude Opus 4.8 Cloud Plan support.
• Added GLM-5.2 and MiniMax M3 Cloud Plan support.
• Added xAI Imagine image generation and provider feature APIs.
• Added xAI video local media upload support.
• Added xAI OAuth credentials, Grok 4.3 model filtering, and improved xAI model visibility.
• Added OpenAI OAuth GPT Image 2 support and improved OpenAI OAuth credential sync and LLM stability.
• Improved Anthropic OAuth account info, account switching, cancel behavior, high reasoning defaults, and image base64 fallback behavior.

Workflows, Hooks, And Developer Tools

• Added a Workflow YAML management editor.
• Added workflow reference checkout support.
• Improved local API descriptions, local API common tool availability, module API schema hygiene, and inline API schemas in generated module skills.
• Added unified tagged logging across Swift, web, and Node.js flows.
• Added a logging clear button.

Capture, Translation, And Polish

• Improved screenshot translation with SSE streaming.
• Added AppShots capture sound controls and exact paired-option hotkeys.
• Improved AppShots context hotkey behavior and live screen attachment sending.
• Improved PopBar row density, keyboard dismissal, and copy shortcut conflict handling.
• Fixed TTS provider voice preview selection, IM channel launch lock recovery, post-tool thinking loading, and collapsed thinking message behavior.

OpenClaw

• Enconvo now supports OpenClaw as an LLM provider — bring any of your OpenClaw agents into Enconvo and run them seamlessly across SmartBar, PopBar, App Sidebar, and every other Enconvo surface

Hermes

• Enconvo now supports Hermes as an LLM provider — every Hermes agent is instantly available across SmartBar, PopBar, App Sidebar, and the rest of Enconvo, so you can use them anywhere you work

File Preview

• Added File preview support — you can now preview files (images, video, audio, Markdown, PDF, and more) right inside SmartBar, beside the chat
• Action buttons on the right side stay visible while previewing
• Fixed an issue with PDF page numbers

MLX

• Local MLX models (beta) — run models on‑device for writing and translation use cases, keeping everything private and fully offline

AI Model Provider

• New provider: Xiaomi MiMo (BYOK & Cloud Plan)

TTS

• New provider: Xiaomi MiMo TTS (BYOK & Cloud Plan)
• New local TTS providers, all powered by MLX and running fully on‑device:
• Kokoro 82M — lightweight and multilingual
• Qwen TTS — multilingual with natural prosody

Dictation

• New on‑device ASR providers, powered by MLX:
• Parakeet ASR
• Whisper ASR
• New realtime / sync ASR providers (BYOK & Cloud Plan):
• Azure (realtime)
• ElevenLabs Scribe v2 (realtime / sync)
• AssemblyAI (realtime / sync)
• Soniox (realtime / sync)
• Volcengine (realtime / sync)
• Your personal Memory is now used as the dictation dictionary, improving recognition of names, jargon, and custom terms you frequently use
• Added Dictation History, with support for using dictation records as Knowledge Base context
• Smoother dictation experience in SmartBar with full realtime / sync mode support

SmartBar

• SmartBar now shows an Update button at the bottom when a new version is available
• Explain follow‑up — keep the conversation going naturally after an explanation
• "Use as context" is now the first action when you use @
• Improved paste image flow and refined screenshot logic for a smoother experience
• You can now delete queued messages while an agent is still running
• Context is now automatically compacted to keep long conversations responsive
• Fixed an issue with @ mention switching

Chat Window

• Message history search lets you quickly find past conversations in the chat window
• Fixed table rendering issues in chat messages
• Fixed: model selection not being applied correctly
• Fixed: model selector dismissing unexpectedly
• Fixed a history switching issue
• Added a Show in Finder icon for delivered files

Settings

• Optimized the shortcut settings page for a smoother configuration experience

Knowledge Base

• Local embeddings — embeddings now run on‑device, so your knowledge base data never leaves your machine
• Knowledge Base entries can now be referenced directly with @ context mentions
• Knowledge Base search now accepts query: string[], enabling multi‑query searches in a single call

IM Channels

• Fixed: IM channels no longer disconnect when an error occurs
• Added a verbose mode — streams the agent's step‑by‑step execution details directly into your IM channel, so you can follow exactly what the agent is doing in real time

Web Search

• New web search provider: Brave Search

Analytics

• Integrated PostHog as the analytics platform — you can opt out at any time from Settings → About

New Logo

• Enconvo has a fresh new look — meet the refreshed app logo

Bug Fixes

• Fixed context handling when returning from another view
• Improved explain context handling
• Fixed links being blocked unexpectedly
• Fixed paste issues in WeChat
• Fixed an issue with the mail notification sound
• Fixed lost context messages
• Fixed an incorrect session directory being used
• Fixed a livescreen image issue

SmartBar

• Resizable SmartBar window — position and size are persisted independently for Command List mode and Chat mode
• New SmartBar suggestion panel
• Dynamic Context: agents now read a live overview of your foreground context and pull in specific items only when needed (for example, ask about the current page directly in your open browser without manually attaching context or opening the App SideBar)
• Smoother task switching with optimized switch logic
• Ask AI: select text inside a SmartBar response and use it as context for follow-up questions
• SmartBar list: menu opens on both click and right-click; added a search box for quickly filtering options
• Pending messages — append new messages while the agent is still running; they are queued and consumed automatically

Dictation

• Refreshed Dictation UI/UX
• Added Qwen ASR support (local, MLX-powered)
• New setting: toggle the start/stop sound effect when dictation is activated
• New setting: toggle whether system audio is muted while dictation is active
• Optimized post-dictation actions — defaults now include Polish and Translate
• New shortcut mode: "Hold or Toggle", default shortcut is Right Command (hold for push-to-talk, tap to toggle)
• AI-powered voice input dictionary for improved recognition of names, jargon, and custom terms

Agent

• New Voice Trigger feature — use Right Option to trigger voice input directly to the agent
• Improved session management
• Optimized context-compression logic

Cron Scheduling

• Cron-scheduled execution for agents — let agents run tasks on a schedule
• Cron logs UI with fixed-height dialog and syntax-highlighted JSON detail view
• Manage cron schedules from both the UI and the API

IM Channels

• Unified IM Channels module supporting Discord and Telegram
• In-channel bot commands: /new, /stop, /voice, /status, /audio
• Access control: pairing-based system with approve/deny notifications

Skills Support

• Skills Store with curated skills (PPT, DOCX, Excel, Obsidian, photo dedup, and more)
• Use skills with the / slash command
• Custom skills directories — pick your own folders or symlinked directories

Knowledge Base

• Reindex ability — rebuild the knowledge base on demand from the UI
• Added xlsx support for knowledge base ingestion
• Audio/video transcription supported in the file loader
• Indexing errors are surfaced in the UI

Memory

• New Memory management UI
• Optimized memory use logic with smarter retrieval and compression strategies

Recording

• Standalone, draggable recording bar with per-screen position persistence

API System

• Expose Enconvo abilities through HTTP API — call Enconvo capabilities from external apps
• Base URL: http://localhost:54535/
• Example:

Enconvo CLI

• Expose Enconvo abilities through a command-line interface
• enconvo -h shows the help message

Computer Use

• New computer-use extension for native macOS control

TTS Improvements

• New providers: xAI TTS, Edge TTS (free), Gemini 3.1 Flash TTS
• New Read Aloud UI/UX: horizontal bar with a live waveform and loading state

AI Model Provider

• New providers: MiniMax, Z.AI (Anthropic-compatible)
• Claude Opus 4.7 (Cloud Plan, Subscription, BYOK)
• Kimi 2.6 / Moonshot updates (Cloud Plan, BYOK)
• DeepSeek V4 Flash and V4 Pro (BYOK and Enconvo Cloud routing)

Image Create

• gpt-image-2 model support (Cloud Plan & BYOK)

Web Fetch

• When fetching x.com, requests are redirected to the Enconvo fetch provider (jina.ai, free tier)

Extensions Improvements

• apple-apps-and-services consolidates all macOS app integrations: Mail, Notes, Calendar, Reminders, Contacts, Messages, and Maps. Replaces the legacy apple_mail and apple_reminders extensions
• media-utils (renamed from video_utils) absorbs audio_utils and adds online video download (YouTube, TikTok, Instagram, etc.) plus audio compress and transcript APIs
• bash util uses RTK to reduce LLM token consumption by 60–90% on common dev commands
• read_file tool now reads images, audio, and video
• image-utils replaces bundled binaries with auto-download from GitHub Releases
• New standalone config module (split out of enconvo) with get / set / preferences endpoints, auto-resolved storeType, and password auto-encryption
• New ai-command extension with migration path from the legacy custom_bot (preserves message history)

UI/UX Improvements

• Refreshed tool-call display
• Tools like write_file now stream their text content live
• Workspace sidebar — quick access to the working directory and deliverables
• Customizable command icons
• Smoother agent flow
• Command settings now jump to a unified settings page

PopBar

• Ask AI integration in PopBar

Other Improvements

• read_file can read image content
• Prompts can reference settings parameters
• Screenshot explanation: clears prior conversation history and no longer auto-copies by default
• Anthropic SDK linked locally with a fork; added jsonrepair for robust streaming parses
• App shutdown: SIGTERM is sent to Python subprocess groups on app termination

Bug Fixes

• Fixed up/down/enter behavior in settings input fields
• Fixed Ollama cloud model issues
• Fixed shortcut handling for menus
• Fixed pasteIn so keyboard events target the last frontmost app
• Fixed reasoning component flicker when content is empty or whitespace

Enconvo in Excel

• Read and edit Excel files directly in the app sidebar, with real-time updates to the currently open Excel file
• Recommended to use Anthropic Claude series models

Claude Opus 4.6 is now live on Enconvo

💡 You can use the new model in the following ways:

• Enconvo Cloud Plan
• Your Claude Pro/Max subscription
• Your own API key

GPT-5.3 Codex is now live on Enconvo

💡 You can use the new model in the following ways:

• Your OpenAI subscription

Meeting/Lectures Recording

• Added meeting recording functionality
• Entry points available in Companion Orb and menu bar
• Records history accessible in Settings
• Can ask questions about recordings
• @Recordings (Meetings/Lectures) allows chatting with all meeting records
• Button to generate meeting summaries (readable version of recording)
• Regenerate transcript support
• Redesigned recording bar to vertical layout
• New transcription model selection UI

Knowledge Base / Memoimization

• Data structuring improvements
• UI logic adjustments
• Batch deletion support
• View detail content
• Note/text types are now editable
• PDF loader defaults to Mistral OCR (supports image recognition, improved content quality)

Ask AI on PopBar

• Select text as context and chat with Mavis in the Smartbar conversation window

Dictation Models UI/UX Optimization

• More intuitive selection and configuration of STT models

Nvidia Parakeet V3

• Added speech recognition model with multi-language support

Hotkey Enhancements

• Support for right-side Command, Option, and Shift keys
• Support for F1–F20 keys in normal mode
• Support for Single Click mode (single-click Command, Option, etc.)

Optimizations

• Chat Window UI/UX Improvements:
• Search chats history
• Rename chat
• optimize delete chat functionality
• Chats now grouped and displayed by date for easier history browsing
• Optimized Markdown Rendering Performance:
• Enhanced markdown rendering efficiency reduce CPU usage
• Full LaTeX formula rendering support
• Code block syntax highlighting support
• Optimized dictation speed and stability
• Sidebar and App Sidebar close behavior now properly resizes application windows
• Improved markdown rendering performance for large text blocks
• Markdown code syntax highlighting
• Sidebar icon improvements (less intrusive)
• Select Area Screenshot now defaults to Mavis bot instead of image compression bot
• rename mini smart bar to companion orb

Bug Fixes

• Fixed Companion Orb positioning issues
• Fixed Smartbar layout issues (1470×956)
• Fixed "Open link in response" issue
• Fixed text highlighting issues
• Fixed messages auto-scrolling to top
• Fixed intermittent message sending failures
• Fixed AI Edit interface display abnormalities
• Fixed Notion MCP installation failure
• Fixed Screenshot Explain/Translation settings page scrolling issue where bottom settings were inaccessible
• Fixed App Sidebar icon not appearing after opening Finder

New Features

• Model Management: Set model properties, delete models, and add custom models

Optimizations

• Optimized credential duplication
• Optimized provider duplication

Bug Fixes

• Improved French language support for getting selected text
• Extension store pagination
• Fixed Nano banana thought_signature issue
• Fixed SwiftData-related crashes
• Fixed LazyVStack crash in SmartBarView
• Fixed issue where eM Client couldn't get selected text
• Fixed AI Edit not getting selected text
• Fixed OpenRouter image generation issue
• Fixed Google account re-authorization issue
• Fixed issue where dropping multiple files from Finder to Smartbar would replace previous files instead of combining them
• Fixed JPG file extension image compatibility
• Fixed Chat Window position and size not saving properly
• Fixed disable app sidebar setting not taking effect
• Fixed Companion Orb visibility and positioning issues
• Fixed SwiftUI-related startup errors
• Fixed issue where launching Companion Orb to Smartbar carried message history

Claude Opus 4.5 is now live on EnConvo

You can use the new model in the following ways:

• EnConvo Cloud Plan
• Your Claude Pro/Max subscription
• Your own API key

New System Toggle Tools

• Toggle System Appearance
• Toggle Full Screen
• Toggle Mute
• Toggle Hidden Files

OpenAI OAuth Additional Models

• gpt-5.1
• gpt-5-codex
• gpt-5.1-codex
• gpt-5.1-codex-max
• gpt-5.1-codex-mini

Bug Fixes

• Fixed UI issue with extension update button
• Fixed issue where user messages sometimes rendered incorrectly

Optimizations

• Chat Window issue fixes and performance improvements
• New Chat / Switch Chat logic optimizations and performance improvements
• Writing Tools now support using the language set in Profile, with multiple language variants (US English, UK English, etc.)
• MiniSmartbar adds screenshot translation
• Smartbar/Small Smartbar are unpinned by default

Bug Fixes

• Fixed issue where clicking elsewhere would trigger the app side bar hint icon menu
• Fixed "App Sidebar icon hide for all" issue

Memory Support

• Add to EnConvo Memory (via right-click menu and dedicated command)
• Use "Add to Memory" in PopBar to quickly add content to EnConvo Memory
• Memory management is available in the Settings page

New Features

• Support for Nano Banana Pro (gemini-3-pro-image-preview)
• Added region-specific language options (American English vs British English)
• Add to EnConvo Knowledge Base (via right-click menu and dedicated command)
• Add "Add to Knowledge Base" to the PopBar

Optimizations

• No AI response needed when using the Knowledge Base tool

Bug Fixes

• Performance issues (especially when using EnConvo in Mail)
• Fixed repeated appearances of the shell runner popup
• Fixed the settings page freezing issue
• Fixed sound when dragging in mail.app on OS26
• Fixed incomplete reading of app content in Safari adaptation

New Features

• Jinja2 syntax is now supported in Prompts
• Quickly select commands in the SmartBar by holding down Command and pressing number keys 1–10

Improvements

• Optimized when the App Sidebar Handle icon is displayed
• Added an App Sidebar toggle and hide option in the right-click menu of the MenuBar and SideBar icons
• Straico now supports image input
• App window content will now include the URL if available
• The SmartBar can now be reset to the global model
• Added a new variable: history_messages

Bug Fixes

• Resolved the issue where the input cursor would automatically jump to the end of a sentence
• Shortened the prompt asking for shell execution permissions
• Improved compatibility for chat window light mode
• Fixed repeated Bun installation issues

Context Optimization

• Support obtaining the content of application windows as context
• Improved interaction settings
• Support selecting available contexts with the # character when inputting text
• You can use @ to specify which tool to use when you input text

Global Sidebar Panel

• A new always-on-top sidebar panel opens on the right side
• The SideBar supports invoking features with @, #, and >
• Default shortcut: Shift + Command + U

Application Sidebar Panel

• Launches an always-on-top sidebar panel on the right within the current application
• Automatically switches to the Bot dedicated to the current Application (such as Finder)
• SideBar supports invoking features with @, #, and >
• Default shortcut: Shift + Command + T

Multitasking

• Multiple tasks can now run simultaneously in the background
• Create a new session with Command + N
• Switch between different tasks by inputting >

Voice Command

• Run a voice command to trigger voice input
• You can specify Window Mode to choose whether the command runs in the Smart Bar or Side Bar
• Set a Default Command to decide which Bot should handle your input after voice transcription
• Default shortcut: Shift + Command + A

My Profile Settings

• You can set your personal description as context to provide to the AI Model
• You can set a global reply language
• Location: Settings > Account > My Profile

Variables (Dynamic Parameters)

• When editing prompts, you can specify dynamic Variables
• Supported Variables are viewable under Settings > Extensions > Variables
• While editing a prompt, activate the Variables list by typing {{ (double curly brace + space)

Custom Bot Creation Optimization

• Improved workflow for creating Custom Bots
• Support for Duplicate Command: Create a new command based on an existing one
• Set reply language independently
• Option to use My Profile or not

Application Support

• Search all installed applications from the SmartBar and press Enter to open
• Every application is an Agent. Use @ to begin a conversation with an application

Screenshot Explain

• Support for taking a screenshot first, then running Explain on it

New Features

• Added Finder right click menu: Add to EnConvo

SmartBar

• Can configure Knowledge base used by Bot on SmartBar
• Support directly enable "Create Video" on the smart bar

Video Generation Provider Added

• Added Kling Video v2.5 Turbo Pro video generation model
• Added Wan 2.5 Preview video generation model

Apple Shortcuts

• Added Apple Shortcuts support, usable in Workflow, Agent, PopBar, SmartBar, etc.

Workflow

• Optimized the logic of deleting edges and nodes in workflow
• Added the ability to reconnect edges in workflow
• Fixed workflow drag variable is undefined

Chat with File

• Optimized file chat (chat mode can directly chat with files)
• Added read document type file tool to agent model (supports pdf, epub, audio, video, word, ppt documents)

New AI Model Provider

• Added Qwen AI Model Provider with API Key

Improvements

• Safety concern for shell runner (alert before execute sensitive commands)
• Optimized Tool Use UI/UX on SmartBar
• Support azure (Microsoft speech) as a provider for live closed captions

Bug Fixes

• Optimized token acquisition for Live Closed Captions
• Fixed screen shot translate bug
• Fixed ollama embedding model not showing up bug
• Optimized message structure (solve the problem of tool use in claude thinking mode)

Web Search Enhancements

• Gemini Search - Use Google Gemini for web search with 2000 free requests per day
• Search UI/UX Optimization - Quickly specify the AI search provider on SmartBar
• Anthropic Built-in Search - Support for Anthropic's built-in web search tool
• OpenAI Built-in Search - Support for OpenAI's built-in web search tool

Image Generation Enhancements

• Image Generation UI/UX Optimization - Quickly specify the image generation provider on SmartBar
• Added fal.ai provider - Support for Seedream 4.0

OCR Enhancements

• AI Model OCR Provider - Choose to use AI models for OCR, supporting all vision models

Improvements

• GPT-5-codex - Use GPT-5-codex for code-related tasks (requires OpenAI OAuth)
• Browser Tab Context - Vivaldi, Wavebox, and other Chromium-based browsers now support getting current tab context

Bug Fixes

• Fixed Screenshot OCR issues
• Optimized Claude authorization
• Improved Qwen OAuth
• Fixed issue where SmartBar opened from PopBar had no actions

Workflow Enhancements

• If Node - Workflows now support conditional logic with If flow control nodes
• Loop-Items Node - Workflows now support looping through a list of items
• File Change Trigger - Added file change monitoring capability to workflows
• Structure Output - Workflow nodes now support structured output format
• Auto Compress Downloads Images - Monitor newly added image files in the download directory and automatically compress them

New Features

• Qwen OAuth Integration - 2,000 free requests per day with no token limits
• Azure Image Generation Support - Added support for Azure OpenAI image generation
• Quit Confirmation - Added Command+Q prompt notification in SmartBar

UI/UX Improvements

• File Settings Item View - Optimized the File Settings Item View interface

Bug Fixes

• Fixed display chaos issues on low-resolution screens
• Fixed Facebook Messenger multiple message copying issue
• Fixed Azure OpenAI error reporting
• Fixed Google Gemini tool use issue
• Fixed input field crash issue

Workflow Enhancements

• Share Workflows via File - Export workflows as JSON files and share them with other users
• Publish Workflows to Extension Store - You can now publish your workflows to the Extension Store

New Workflow Templates

• Sync to Apple Notes - Sync content to Apple Notes
• Nano Banana + Hailuo Video Generation - Workflow based on Nano Banana and Hailuo Video Generation

fal.ai Supports EnConvo Cloud

• Hailuo 02 Standard/Pro Text-to-Video (Free: 5 uses per 24 hours)
• Hailuo 02 Standard/Pro Image-to-Video (Free: 5 uses per 24 hours)
• Veo3 Fast/Standard and Image-to-Video
• Seedance V1 Lite Text-to-Video

Optimized Features

• Edit Command Description - Edit command descriptions for better Agent utilization

Bug Fixes

• Fixed issue where workflow links sometimes become invalid
• Disabled commands/bots are showing up

New Features

• ChatGPT Plus/Pro in EnConvo - Use your ChatGPT Plus/Pro subscription in EnConvo, no API key needed

New Models

• Nano Banana (Gemini 2.5 Flash Image) - Google DeepMind's Gemini 2.5 Flash model with image generation and editing capabilities

Optimized Features

• Image Compression - Images are now compressed before being sent to models
• Node Performance Optimization - Automatically clean up idle workers
• Youtube Downloader Update - Upgraded yt-dlp to version 2025.08.22

Bug Fixes

• Fixed incorrect path handling issue in video compression
• Fixed Terminal.app not syncing issue
• Fixed AI model global display issue

New Features

• Claude Pro/Max in EnConvo - Use your Claude Pro/Max subscription in EnConvo, no API key needed

Optimized Features

• Thinking Effort - Quickly configure the model's deep thinking capability (Command + E)
• MCP OAuth - Use MCP services that support OAuth, such as Notion (Official)

New Free Services

• Voyage-3.5-lite (embedding) - Powerful multilingual embedding capabilities for free

UI/UX Improvements

• Custom bots - Added New Bot button for quickly creating custom bots
• Model selection - Model selection is now displayed in a double-row layout

Others

• Allow uninstallation of some non-essential pre-installed Extensions

UI/UX Improvements

• Model Card - Show detailed model information and parameters when hover on the model name

Optimized Features

• Global Model Provider Settings - Switch the model provider between global settings and custom settings
• Add/Edit Model - Adjust the model's parameters based on actual attributes and preferences

New TTS Models

• MiniMax Speech 2.5 Model - Enhanced Multilingual Expressiveness Exceptional Voice

New MCP Server

• Excel MCP Server - Manipulate Excel files without needing Microsoft Excel installed

Brand New Extension Store

• One-click installation - Say goodbye to tedious configuration
• Multiple Registries - Choose from EnConvo official, Smithery, or mcphub extension sources
• DXT One-Click Installation - Quickly install plugins from dxt files
• Extension Upgrade - Independent extension upgrades with one-click functionality
• Manual Installation - Support for Stdio/Http/Sse type MCP extensions

New AI Models (Limited Time Free, Ending on September 30th)

• OpenAI gpt-oss-120b, gpt-oss-20b

Link Reader Providers

• Local Fetch - (locally, free)
• Crawl4AI - (locally, free)

Website Crawler Providers

• Crawl4AI - (locally, free)

New Features

• Points Usage Log - Added points usage log page (Settings->Account->Points Usage)

Optimized Features

• MCP Server Auto-shutdown - MCP servers automatically shut down after 3 minutes of inactivity

Added Speech Recognition Model

• mlx parakeet-tdt-0.6b-v2 - Faster speed and better quality, local & privacy-focused

New AI Model Provider

• 302.ai - Added 302.ai LLM provider

Optimized Features

• App Size Optimization - Reduced application size from 367MB to 224MB
• Moonshot BaseURL Support - Added support for custom BaseURL settings

Bug Fixes

• Fixed Online Video Downloader audio only option and mp4 suffix issues
• Fixed MCP runtime error issues
• Fixed Summarize YouTube/Link response language setting not taking effect
• Resolved fireworks.ai error issues

Knowledge Base

• Enhanced Knowledge Base UI/UX - Streamlined interface for easier management
• YouTube Video Import - Import videos with automatic subtitle extraction
• Image Import with OCR - Extract and index text content from images automatically
• Audio File Import - Convert speech to text and add to knowledge base
• Video File Import - Extract audio and convert speech to searchable text
• Single Web Page Import - Extract content from any webpage
• Whole Website Import - Crawl and import entire websites using Firecrawl
• Direct Text Addition - Quickly add custom text content

New AI Models

• Grok 4 - Added support for Grok 4 (EnConvo Cloud Plan and API key)
• Kimi K2 - Added support for Kimi K2 (EnConvo Cloud Plan and API key)

TTS

• TTS Voice Preview - Support for previewing TTS Voice sounds

New Features

• Firecrawl - Added Firecrawl as Link Reader Provider
• Website Crawl - Added Website Crawl provider support
• Translator (Interactive) - Added Translator (Interactive)

Optimized Features

• Summarize Url / Chat with Url - Optimized continuation conversation logic
• YouTube Video Download - Optimized download and subtitle download logic
• Chat with Youtube - Optimized logic and improved prompt
• ElevenLabs V3 - Support for ElevenLabs V3

Bug Fixes

• Fixed copy/paste compatibility for Ukrainian language
• Fixed video playback issues for downloaded videos

Mini SmartBar

• Mini SmartBar now supports custom commands with Application Awareness and Browser Tab URL Awareness

Dictation

• Added Dictation Post Actions feature (Fix Spelling, Translate, customized workflow)
• Renamed Voice Input to Dictation

Prompt Management

• Added Prompt Store for installing curated prompts
• Added Installed Prompt management
• Added Create new prompt command

SmartBar

• Added SmartBar Navigation bindings (Control+P/N or vim-like Control+J/K)
• Added SmartBar Auto clear conversation

New Commands

• French Teacher: A French teacher to help users learn French

Optimized Features

• Disable command feature: disabled commands won't appear in SmartBar
• PopBar now supports detecting current browser tab URL
• Shortcut settings UI/UX optimization

Bug Fixes

• Fixed issue where Dictation Bar couldn't be closed
• Fixed issue where PopBar window position exceeded screen boundaries
• Fixed Shortcut settings invalidating existing shortcuts
• Fixed Post Action not working in PopBar

PopBar (Big Update)

• Application-Awareness: PopBar can detect the current application and display different toolbars
• Built-in Writing Tools PopBar Instance for writing apps
• Added option to automatically replace current text after tool returns result
• Optimized PopBar appearance timing and position
• Added Open Link tool, Search Google tool
• Added PopBar disable in this application feature
• Added Duplicate of PopBar Instance feature
• Added Modifier flags action feature

New Commands

• Added more writing tools: proofread, rewrite, changeToneToProfessional, makeConcise, extractKeyPoints, convertToList, convertToTable
• Added AI Editor command (similar to Cursor.app's Quick Edit)

New Extension (Fal.ai)

• Generate video with Veo3
• Fal.ai text-to-video with Minimax Hailuo02 Standard/Pro
• Fal.ai generate video with Bytedance Seedance 1.0 Lite

Optimized Features

• Added custom bots management in settings
• Adjusted Mini SmartBar expansion timing
• Added PopBar and MiniSmartBar toggles in menu bar
• Added delete command and reset configuration features
• Optimized Command search UI/UX

EnConvo Cloud Plan New Models

• Claude Sonnet 4, Opus 4
• O3, O3 Mini, O4-Mini
• Gemini 2.5 Pro, Gemini 2.5 Flash, Gemini 2.5 Flash Lite Preview

Workflow

• Added Generate Video with Veo3 workflow template
• Added Text to Audio workflow template

New Features

• Configurable option for whether new chats in Chat window will be displayed in Smartbar

Optimized Features

• Optimized Installed Extensions management page
• Optimized Explain prompt
• Optimized PopBar settings page UI/UX
• New settings UI/UX for Smartbar, MiniSmartbar, ChatWindow, Langchain, Developer Tools

Bug Fixes

• Fixed Create New bot editing issue
• Fixed Google Cloud Search errors
• Fixed chatbot name change issue
• Fixed model list refresh issue

New Features

• Added scheduled trigger functionality for Workflows with cron expressions support
• Pre-built workflow template for daily weather information with TTS and email

Optimized Features

• Workflow/Agent supports Gmail attachment sending
• Ability to rename nodes in Workflows
• Optimized auto-scroll issue when outputting text in chat

Fixed Issues

• Straico AI issues fix
• Fixed Shortcuts Center display issue
• Fixed Google Gemini error issues
• Fixed clipboard-related crash issues

Workflow

• Brand new visual workflow editor
• Built-in 200+ tools, native support for MCP Servers in workflows
• Smart parameters support - automatically generate tool parameters through AI
• New Workflow Templates Store
• Workflow parameter settings support jinja2 template syntax

Brand new Credentials Management

• Optimized Credentials management logic
• Support for OAuth 2.0 authentication
• Added more Provider Credentials management

New Plugins

• New Gmail with OAuth2 authorization
• New Firecrawl (including EnConvo Cloud support)
• New Gemini TTS, Gemini TTS Multi Speaker

MCP Feature Improvements

• MCP can now individually configure whether to enable specific tools
• MCP is now bound to Agent, no longer globally applied
• MCP management add/delete UI/UX improvements

Context Awareness Feature Improvements

• Support for more Context awareness (screenshots, clipboard text, etc.)
• Users can freely configure the context that each Command can perceive

Feature Improvements

• Fix Spelling&Grammar supports text highlight comparison
• Stability and performance optimizations
• Settings page UI/UX improvements
• Chat Window new session creation UI/UX improvements
• Mini smartbar automatically shrinks and hides after 1s of no operation
• Mini smartbar automatically snaps to edges
• Create New Bot UI/UX improvements

Improvements

• Language Settings for Chat Responses - Configure the language your agent responds in
• Toggle for Mini Window - Enable or disable the Mini Window feature
• Enhanced Voice Dictation Stability - More reliable speech recognition

New Features

• Introduced Mini SmartBar - A compact desktop companion with quick access to features
• Integrated Hailuo Voice as default TTS Provider (EnConvo Cloud Plan)
• New OCR Providers
• Apple Native OCR: Fast, offline, privacy-focused local processing
• Mistral OCR: Integrated solution supporting EnConvo Cloud Plan and custom API Keys
• Added new Mistral models to EnConvo Cloud Plan
• Mistral Large (300 points per message)
• Mistral Small (Free)
• Introduced EnConvo Cloud Plan for Speech-to-Text Provider
• Supporting whisper turbo, gpt-4o-transcribe, and gpt-4o-mini-transcribe

Improvements

• Startup optimization and bug fixes
• Enhanced Image Generation with Flux1.1 pro
• Added close button to Voice Input Method
• Added direct model configuration in Chat Window
• Expanded Google Gemini model support
• OCR Improved: Automatic language detection with Japanese support
• Optimized audio auto play logic
• Enhanced Speech-to-Text stability

New Features

• Introduced MCP Store - Your One-Stop Shop for MCP Servers
• One-click installation for quick and easy setup
• Online access at https://app.enconvo.ai/mcp_store
• Share your favorite MCP Servers with friends via shareable links
• Added dedicated MCP configuration toggle in SmartBar
• Integrated support for gemini-2.5-pro-preview-03-25 (EnConvo Cloud Plan)
• Introduced Command + S shortcut to toggle AI search

New Features

• Introduced Mavis - The general AI Agent for your Mac
• Vibe Coding: Generate clickable project links and ready-to-run applications
• Report Writing: Create detailed, clickable reports from simple descriptions
• Added multi-language UI support (German, French, Japanese, Spanish)
• Introduced Text to Sound Effect feature powered by ElevenLabs

Improvements

• Rewrote message list for enhanced loading and scrolling performance
• Added resolution settings for Online Video downloads
• Enhanced video downloader with browser cookie support for private content
• Improved knowledge base addition speed and stability

New Features

• Added support for new Perplexity models: r1-1776 and sonar-deep-research
• Added configurable result count for AI Web Search
• Added ability to hide Chat Window using the ESC key
• Added voice playback functionality for existing responses
• Added Elevenlabs Scribe speech-to-text integration
• Added Elevenlabs TTS and Speechify TTS integration

Improvements

• Optimized Live Captions with customizable caption height
• Enhanced stability when adding large files to knowledge base

Bug Fixes

• Fixed longPress shortcut key issue with Voice input method push

New Features

• Added SmartBar default command setting functionality
• Added reasoning_effort parameter for LLM

Improvements

• Optimized PopBar hotkey logic to toggle display/hide mode
• Enhanced SmartBar UX logic

Bug Fixes

• Fixed issue where "New Chat" couldn't use web browsing functionality
• Fixed SmartBar hotkey deletion not working

New Features

• Added support for Claude 3.7 Sonnet and Claude 3.7 Sonnet Thinking models
• New Local Link Reader provider for faster, more stable webpage content extraction
• New Read Aloud Webpage Tool

Improvements

• Optimized prompts to improve agent instruction following capabilities
• Enhanced MCP (Model Context Protocol) stability

New Features

• New Online Video Downloader tool (TikTok, YouTube, Twitter, Instagram, Reddit, Vimeo, etc.)
• New App Context Awareness feature
• New post-action settings (paste, insert, copy)
• New PopBar display style settings

Improvements

• Enhanced AI Web Search with source references
• Improved Chat Window's new chat and clear conversations logic
• Upgraded Speech-to-text Deepgram Provider with Nova 3 model support
• Optimized DeepSeek R1 model speed and stability
• Added support for custom Google Gemini and DeepSeek LLM Provider BaseUrl

Bug Fixes

• Fixed auto duplicate issue in Preview.app
• Fixed text input issues with Voice Input Method
• Fixed Claude image format compatibility issues
• Fixed Insert Below compatibility issues
• Fixed model display not updating after switching models

New Events

• DeepSeek R1 and DeepSeek V3 are now available for unlimited free use (2025-02-13 ~ 2025-03-13)

Improvements

• Optimized default configuration settings logic

Bug Fixes

• Fixed shortcut key settings not taking effect
• Fixed issue where models could not be selected via mouse click

New Features

• Added model switching with / character
• Added Context deletion features
• Deep Research Agent now shows research process
• Added custom shortcut keys support for @ and / trigger characters

Improvements

• Improved startup stability
• Enhanced Context display
• Improved copy/paste compatibility

New Features

• Added Recraft integration for enhanced design workflows
• Added Image Editor Agent for advanced image manipulation
• Added New Chat functionality
• Added Duplicate Command functionality
• Recraft tools: Crisp Upscale, Generate Image, Image to Image, Vectorize Image, Creative Upscale, Remove Background, Replace Background

Bug Fixes

• Fixed Exa Websearch Provider issues
• Fixed incorrect tab detection in Safari

New Features

• Added Deep Research Agent

Feature Updates

• Changed default LLM model to gpt-4o

Bug Fixes

• Fixed background crash issues
• Added multi-screen centered display support for Smart Bar

Google Gemini Updates

• Updated Gemini 2.0 Flash to stable release
• Added gemini-2.0-flash-lite model
• Added gemini-2.0-pro-exp model

Subscription Management

• Added subscription cancellation management interface

Bug Fixes

• Fixed Hailuo TTS (Chinese) related issues
• Fixed missing text issue when activating PopBar with hotkey
• Fixed Double Click hotkey configuration issue
• Fixed Voice Input Method unable to stop when connection is lost

Video Downloaders

• Twitter Video Downloader: Added support for downloading Twitter videos
• YouTube Video Downloader: Added support for downloading YouTube videos

Apple Extensions Enhancements

Apple Reminders

• Update reminders, Delete reminders, Toggle reminder status, Get all incomplete reminders
• Delete events, Update events, Get calendar events list
• View unread emails, Delete emails, Toggle read/unread status, View email content, Send emails

Screenshot Extension Updates

• Capture current application, Capture current screen, Manual window selection capture

AI Model Improvements

• Optimized thinking process display for Ollama R1 and other reasoning models
• Added DeepSeek R1 and DeepSeek V3 model support to EnConvo Cloud Plan
• Enhanced Jarvis prompt for improved Agent task execution capabilities
• Updated Perplexity models: sonar, sonar-pro, sonar-reasoning

Other Updates

• Conch TTS: Split into Chinese and international servers
• LM Studio: Added model list display
• Fixed straico-related issues

Agent Features

• Jarvis: Now automatically completes tasks using tools based on your needs
• Available tools include: AppleScript, Shell, Python, NodeJS, Image Generation, Apple Calendar, Apple Reminders, Image Compression, HTTP Requests, TTS, Audio/Video Transcription, Video Compression, Web Search, Link Reader, YouTube Loader, File System Operations
• Added "Create New Agent" feature

MCP (Model Context Protocol) Support

• EnConvo as MCP Client supports installation and use of any MCP Server
• MCP Settings page with real-time server status monitoring

DeepSeek r1 Model Support

• Added support for DeepSeek r1 model
• Added Thinking interaction mode

New Plugins

• Video Compress, Extract Audio, Audio Compress, Transcribe Audio/Video Files
• Python Code Runner, NodeJS Code Runner, Shell Script Executor, AppleScript Executor
• Conch TTS Support

Google LLM Provider Updates

• Updated to latest gemini-2.0-flash-thinking-exp-01-21
• Native support for audio message types

Optimizations

• Improved TTS performance
• Optimized Python environment initialization
• Added parameter display in Workflows

Chat Window Enhancements

• New UI design for improved aesthetics
• Performance optimization with reduced CPU usage
• Expanded functionality: Voice Input, Screenshot, Web Search, Live Screen, Live Camera, File Attachment, and Realtime Call

YouTube Plugin Integration

• Chat With YouTube, Summarize YouTube, YouTube Transcript Loader, YouTube Context Awareness

Code Runner Plugin

• Run shell, Node.js, Python, and AppleScript code

Straico Integration

• Straico LLM Provider, Image Generate Provider, and Plugin (Chat, RAG, Agent)

Google Gemini Update

• gemini-2.0-flash-exp, gemini-2.0-flash-thinking-exp-1219, gemini-exp-1206

X.AI-GROK Update

• grok-2-vision-1212, grok-2-1212

New Setting Option

• Added "Context Message Count Limit" setting

Performance Improvements

• Voice Input Efficiency: CPU usage reduced from 40% to 5%
• SmartBar Optimization: CPU usage reduced from 1.9% to 0.2%
• Live Screen Preview Enhancement: CPU usage reduced from 30% to 12%

New Features

• Ollama Embedding Integration: Chat with Document and Knowledge Base features now work completely offline
• Audio Input Selection: Voice Input Method now supports custom audio source selection

Improvements

• UI Enhancements: Fixed "Untitled" display issue, resolved text styling in SmartBar, fixed pinyin input display issue
• Voice and TTS Improvements: Fixed TTS speed settings, resolved incomplete voice recognition

Integrations

• Google TTS Support

New Features

• Deeplink Support: Generate deeplink for command in SmartBar for quick access
• Markdown-to-Image Extension: Convert Markdown content into beautiful images
• EnConvo Companion Chrome Extension: Directly capture and access current browser content
• ChatGPT Website Integration: Quick access to ChatGPT directly from SmartBar
• Claude Website Integration: Launch Claude instantly from SmartBar

Improvements

• Context Awareness: Added ARC browser support
• Voice Input Method Enhancements: Stream Insert mode toggle, 2-second pause detection for automatic line breaks
• Other Improvements: Reduced default keyboard shortcuts, Slovak language support, new extension removal feature

Bug Fixes

• Fixed workflow selection issue during creation
• Resolved "Untitled" workflow display problem in PopBar