Every fix, listed.
A detailed, customer-readable record of 586 fixes across the latest release ranges. Updated 19 September 2026.
The current version is v0.3.255, released 19 September 2026.The Updates page has a short card for each release: what is new, what improved, and the headline changes worth knowing about, newest first. Updates →Release
Fixes in v0.3.255
282 fixes across 10 categories.
Telephone calls
20 fixes- v0.3.185: Player One can answer and place telephone calls, with the whole conversation kept in the chat.
- v0.3.185: A live banner sits above the composer while a call runs, with a button to hang up.
- v0.3.185: Only the part of a reply meant to be spoken is read aloud, and the rest stays in the chat.
- v0.3.185: A caller's words appear as an ordinary message, marked as spoken.
- v0.3.185: A phone line keeps its credentials in the vault, never in a visible record.
- v0.3.185: A call no longer breaks off in silence when the caller interrupts.
- v0.3.185: The call card reads all six ways a call can end.
- v0.3.186: A caller reaches a chat that already exists for them, so the agent remembers the last call.
- v0.3.186: An abandoned call leaves no half-finished record behind.
- v0.3.186: A call the agent ends politely is recorded as a goodbye, not a refusal.
- v0.3.186: A phone line can be changed in place without losing its number.
- v0.3.187: An agent on a call speaks instead of typing.
- v0.3.187: A sandboxed agent can speak on a live call.
- v0.3.187: A call started from the owner's chat runs with that chat's permissions.
- v0.3.187: Numbers are read in international format, so a caller is matched however they dial.
- v0.3.187: The Phone lines admin page lists your lines with their recent calls beneath.
- v0.3.187: A call that refuses to end is retried and then closed.
- v0.3.187: The ended card says why a call finished, in the provider's own words.
- v0.3.187: Messaging and Webhooks moved into a Communications group in Admin.
- v0.3.192: A phone call uses the same model and thinking level as any other chat.
Your own computer
39 fixes- v0.3.196: Player One works on your own computer through Sidekick.
- v0.3.196: Commands run on your computer in a shell that stays where you left it.
- v0.3.197: You can sign in to Sidekick through your browser.
- v0.3.197: Player One can list the computers you have paired, and forget one.
- v0.3.197: Output from your computer is shown, kept and paged under three separate limits.
- v0.3.198: Player One can capture your screen, and see your displays and windows.
- v0.3.198: Player One can act on your computer, clicking, typing and dragging.
- v0.3.198: Player One can work with the files on your computer.
- v0.3.198: A change of access mode appears in the chat as a single note.
- v0.3.198: The chat says when a paired computer disconnects, and names it when something fails.
- v0.3.198: A long-running command tells you the chat carries on by itself.
- v0.3.198: A request with no computer attached refuses by name instead of failing obscurely.
- v0.3.199: A computer that cannot do something says so, rather than appearing to be absent.
- v0.3.205: Four access modes decide how much a paired computer allows.
- v0.3.205: The chat shows what a computer is allowed to do beneath its mode.
- v0.3.205: A refusal names the mode you actually see in the picker.
- v0.3.222: Sidekick has a side panel on your desktop, with its own chat.
- v0.3.222: A preview opens over the side panel instead of sending you to a browser.
- v0.3.222: The side panel's header shows the agents in the chat, and drags the window.
- v0.3.223: A signed-in Sidekick stays signed in.
- v0.3.229: A refusal from your computer keeps its own reason, and an installation problem is named as one.
- v0.3.231: The side panel lists your chats from every project, in folders, loading more as you scroll.
- v0.3.231: The side panel stays on its own page, wherever the site navigates.
- v0.3.231: The panel's header carries the other agents, and a close button.
- v0.3.231: The panel's composer is the full desktop composer, at desktop size.
- v0.3.231: Menus and dialogs inside the panel read at the panel's own size.
- v0.3.234: A page preview opens a viewer window in the side panel, so a browser session is visible at all.
- v0.3.237: The panel's chat list is the same list as the sidebar's.
- v0.3.237: Menus, submenus and popovers stay inside the panel's column.
- v0.3.237: The panel's composer takes the desktop controls in a flat frame.
- v0.3.237: A mark in the panel's header starts a new chat, and says why one waits under All projects.
- v0.3.238: The panel signs in through the extension that frames it.
- v0.3.238: The parent avatar in the panel opens the parent chat.
- v0.3.242: A change to what a paired computer allows takes effect while it stays connected.
- v0.3.245: A face in the agent strip opens that chat inside the Sidekick panel.
- v0.3.248: An action on your computer checks itself and reports the new state in one step.
- v0.3.251: A command on a paired computer reads as finished when it finishes.
- v0.3.251: Typed text keeps its edges, and a refusal no longer suggests the thing it just refused.
- v0.3.255: A text area on a Mac can be reached by its role alone.
Chat and conversations
54 fixes- v0.3.177: The compaction notice offers to start a new chat.
- v0.3.183: Opening the side preview no longer moves you earlier in the thread.
- v0.3.183: A Preview control opens the workspace pane.
- v0.3.186: A new chat opens on a redesigned welcome screen.
- v0.3.186: The compaction marker sits at the end of the transcript.
- v0.3.187: Your own passthrough chats stay in the sidebar and in Recent chats.
- v0.3.189: A live progress line replaces the activity boxes.
- v0.3.189: The progress line says when the connection drops.
- v0.3.189: A prompt card reads the project you are in, not one selected elsewhere.
- v0.3.192: A light rule separates the messages of one reply.
- v0.3.192: A finished answer beneath a run still going reads as in progress.
- v0.3.194: The mobile Photo button keeps the photo it was given.
- v0.3.195: No sideways scrollbar when the window is too narrow for avatars.
- v0.3.196: The progress line opens into a log of what is happening.
- v0.3.196: Steps are coloured by kind, so tools and skills are told apart at a glance.
- v0.3.197: A yielded command no longer splits the reply into separate cards.
- v0.3.197: The progress line names what a tool is acting on.
- v0.3.197: Queued messages arrive one at a time instead of as a single block.
- v0.3.217: A shortened command result says exactly where to continue reading.
- v0.3.219: A helper waiting on your approval tells the agent it works for straight away.
- v0.3.224: Agent faces sit over the composer, replacing the cards that were above it.
- v0.3.224: A chevron separates the parent chat from the agents it started.
- v0.3.226: A keyboard shortcut starts and stops a voice note.
- v0.3.226: Voice options open in a wide panel, and recording pauses while it is open.
- v0.3.227: The composer's effort chip shows the level the next turn would actually run at.
- v0.3.227: A narrow window keeps the input focused after you send.
- v0.3.229: Fold and email notices read as lines in the chat rather than messages from an agent.
- v0.3.230: A timer's report reaches the channel it was set from.
- v0.3.231: An agent working in the background wears a turning ring on its face.
- v0.3.231: A chevron marks the parent chat in the row of agent faces.
- v0.3.231: The agent faces above the composer sit better, and lift as you point at them.
- v0.3.231: A voice note shows a moving level meter while you record.
- v0.3.236: The chats an agent started are grouped under it in the sidebar, as a tree you can open.
- v0.3.236: A cog on each chat carries its options in one place.
- v0.3.236: The sidebar filters by what a chat is, so scheduled and flow runs can be listed on their own.
- v0.3.238: Pickers open as bottom sheets on a phone, and in the side panel.
- v0.3.238: The model, effort and agent pickers show a card as you point at a choice, and the model picker ticks the one in use.
- v0.3.238: Keyboard focus shows a soft ring instead of the browser's black outline.
- v0.3.238: The row of agent faces no longer reserves empty space above the composer.
- v0.3.206: An expired approval says so, and a late click still gets an answer.
- v0.3.207: Messages sent while a chat is opening wait for it instead of being lost.
- v0.3.209: An approval card's command scrolls on a phone instead of being cut off.
- v0.3.209: An approval card hides values that look like secrets, and leaves the rest readable.
- v0.3.239: A reply that folded several model calls says so, and names its last prompt.
- v0.3.240: Compaction reads as one card, from the request through to the result.
- v0.3.240: The stop button is disabled while a conversation is being compacted.
- v0.3.241: A new chat's effort chip reads the route rather than a default.
- v0.3.243: Tool result panes show their size.
- v0.3.246: The notice shown while an installation is away says what it usually means.
- v0.3.248: A screenshot already shown stops re-appearing as a long reply grows.
- v0.3.248: A screenshot step keeps its tool call and its thinking once the picture is attached.
- v0.3.248: Streamed text keeps pace with a fast model instead of landing all at once.
- v0.3.254: Menu and popup text stays sharp, with ClearType kept on Windows.
- v0.3.254: A selected helper in the sidebar keeps its whole level in view.
Jobs, tasks and automations
25 fixes- v0.3.177: A coordinator wakes as soon as a worker finishes.
- v0.3.177: Creating a job refuses an invalid workspace straight away.
- v0.3.183: Stop, resume and close a job from the job page.
- v0.3.183: The job board's third column says Finished.
- v0.3.183: Stopping a job names the person who stopped it.
- v0.3.183: Scheduled tasks pause and resume, rather than being enabled and disabled.
- v0.3.183: Flow runs survive a restart instead of being discarded.
- v0.3.183: A resumed flow that makes no progress stops instead of spinning.
- v0.3.189: A throwaway flow's last build can be removed.
- v0.3.205: A scheduled run takes its permissions from the task's owner.
- v0.3.206: A stopped worker cannot overwrite its own record.
- v0.3.207: A schedule that cannot happen is refused as you enter it.
- v0.3.207: An impossible schedule reads as never, not as every two seconds.
- v0.3.217: A scheduled task keeps the thinking level you chose for its chat.
- v0.3.217: A scheduled task that names a model your installation does not allow fails by name.
- v0.3.207: A scheduled run does its work in the project its task belongs to.
- v0.3.209: Removing a schedule takes its project membership with it.
- v0.3.221: A job in which no task was ever done no longer reads as complete.
- v0.3.221: A failed task parks its job instead of leaving it adrift.
- v0.3.221: Finishing a run names the runs still keeping a job open.
- v0.3.228: A job with live work or an open review is not cut short by the resume cap.
- v0.3.230: A refused worker request is recorded rather than passing in silence.
- v0.3.231: A job can be given its own limit on how many times it retries.
- v0.3.231: A parked job is flagged once, not on every check.
- v0.3.241: A deleted job's queued announcements go with it.
Agents, skills and tools
21 fixes- v0.3.177: Changing a chat's project reloads the skills for that project.
- v0.3.183: A skill's notes are shown on its detail page.
- v0.3.183: The Skills page says where a skill came from.
- v0.3.183: The Skills page no longer claims a control it does not have.
- v0.3.185: Searching skills stops re-reading the whole workspace each time.
- v0.3.186: Ten shipped skills whose details never loaded now work.
- v0.3.187: A stale tool can be removed from one agent, or from every agent at once.
- v0.3.187: An agent's card says when it is sandboxed.
- v0.3.187: Deleting an agent offers to delete its Memory Space with it.
- v0.3.187: A skill's emoji shows in search results.
- v0.3.189: A permission entry that matches nothing explains why.
- v0.3.194: A turn that resumes after an approval keeps its tools.
- v0.3.196: A project skill is created with the visibility you asked for.
- v0.3.216: Every agent can read the model catalogue and switch its own session.
- v0.3.216: A worker runs on the model named by whoever started it.
- v0.3.244: A map of the helpers working on a chat, opened from the row of agent faces.
- v0.3.244: The helper tree shows every level beneath the chat you are in.
- v0.3.244: A helper that is working wears a turning ring on the map, as it does in the strip.
- v0.3.244: Helper popups and sidebar rows lead with the chat's own name.
- v0.3.245: The sidebar keeps the parent's row and its helper list in view while a helper is open.
- v0.3.245: A helper opens like any other chat.
Memory, knowledge and search
13 fixes- v0.3.183: A healthy search index is no longer reported as failed.
- v0.3.183: Search never drops a source you asked for without saying so.
- v0.3.185: An agent searches its own memory and only its own.
- v0.3.185: Two files are two results, however similar their openings.
- v0.3.186: This installation's own Fold projects can be connected as knowledge sources.
- v0.3.186: A knowledge source on our own service needs no key of your own.
- v0.3.187: A knowledge source whose project has gone is marked, skipped and removable.
- v0.3.187: Removing a Memory Space asks whether its search index goes too.
- v0.3.187: A scan finds a Memory Space wherever it lives.
- v0.3.226: A knowledge source is checked with the credential that can actually see it.
- v0.3.229: Searching for a skill and loading it now agree on what exists.
- v0.3.230: A run that ends unverified parks only its own Memory Space.
- v0.3.231: A Memory Space still being indexed reads as indexing, not as failed.
Models and providers
28 fixes- v0.3.177: Routing explains the decisive reason it fell back to another model.
- v0.3.177: A billing refusal is shown as billing, and keeps the provider's own remedy.
- v0.3.182: Provider tabs follow the order you set.
- v0.3.182: The status panel says when it last checked, and offers to look again.
- v0.3.182: The status panel stops blinking, and reads at a normal size.
- v0.3.186: Each model in a routing chain gets its own thinking level.
- v0.3.186: A missing model is reported as missing, not as an unreadable error.
- v0.3.187: The effort controls line up down a chain.
- v0.3.189: A provider you switched off stops being offered.
- v0.3.189: A connected xAI account shows its Grok Imagine models.
- v0.3.216: New models this release: Gemini 3.8 Flash, GPT-6 Astra, GLM-5.3, GLM-5.3 Flash, Qwen3.8-Flash-Next and Fable 5.1.
- v0.3.217: A provider failover does not overwrite the model you pinned.
- v0.3.218: A model refusal returns as a clear refusal rather than an unexplained failed run.
- v0.3.221: An unknown thinking level is refused plainly.
- v0.3.227: New model this release: DeepSeek V4.1 Flash.
- v0.3.227: The retired DeepSeek V3 names are gone from the catalogue.
- v0.3.234: A run that falls back to another model uses the right thinking level for the model actually serving it.
- v0.3.236: An Anthropic account on extra usage credits is no longer reported as out of usage.
- v0.3.236: A refusal about your plan names the ChatGPT subscription rather than Codex.
- v0.3.237: A model one account's plan lacks is tried on your other accounts.
- v0.3.237: The signed-out notice names the provider and the account.
- v0.3.238: Retired models no longer linger in the catalogue after an update.
- v0.3.246: On Anthropic models the tools sent with every request sit in one stable cached block, so a long conversation costs less and starts faster.
- v0.3.246: Loading a tool part-way through a conversation no longer rebuilds that cached block.
- v0.3.240: A substituted model resolves its own thinking level.
- v0.3.248: You confirm a provider's own data terms before that provider is connected.
- v0.3.248: TypeSafe can be connected, with a vendor-neutral tool for judging output.
- v0.3.248: A Grok chat is no longer refused when the judging tool is present.
Admin, settings and security
29 fixes- v0.3.183: Gateway is now called Runtime, and Instance is now called Installation, throughout the interface.
- v0.3.183: The Settings page stops having three names for the same thing.
- v0.3.183: You can test a connection from the page where you added it.
- v0.3.183: The help panel is called Help.
- v0.3.183: Plain words on the Backups import tab.
- v0.3.183: API keys read the token the runtime actually sends.
- v0.3.183: A hosted link is handed over with a privacy notice.
- v0.3.185: An approval message never shows the raw command.
- v0.3.187: A command recorded in the chat never shows its secrets.
- v0.3.196: Memory pressure is measured against one ceiling, and notices name it.
- v0.3.201: A vault refusal names the field and the shape it wanted.
- v0.3.217: A local settings fault ends the run without wrongly blaming a provider.
- v0.3.217: The default agent cannot be accidentally made sandboxed.
- v0.3.206: Reading another program's memory needs a grant you give.
- v0.3.206: Every command runs as the ordinary runtime user.
- v0.3.207: Deleting a vault item names what still uses it, and one action clears them.
- v0.3.209: A stored permission entry never keeps the raw command line.
- v0.3.235: Prompt capture is a debug switch that writes outside your workspace, and is off unless you turn it on.
- v0.3.235: A one-time sweep takes old prompt captures out of customers' workspaces.
- v0.3.235: Prompt captures never cross a workspace sync, and never reach cloud storage.
- v0.3.235: A Prompt capture switch sits behind Advanced in Settings.
- v0.3.236: The tool catalogue reports what each permission profile grants.
- v0.3.236: The deny picker offers every tool an agent can actually call.
- v0.3.236: A tool configuration update keeps the overrides you set.
- v0.3.238: The Admin menu opens with a Back to site row.
- v0.3.238: A vault injection the runtime does not recognise is named rather than silently dropped.
- v0.3.252: A command approval can be answered for five minutes rather than two.
- v0.3.252: An Always allow answer sits on the approval card and in the operator alerts.
- v0.3.253: An Always allow answer never grants an interpreter.
Files, backups and media
18 fixes- v0.3.177: Dropbox team folders that could not be reached now sync.
- v0.3.183: The agent browser opens maximised.
- v0.3.186: The browser has a memory limit and a time limit for page actions.
- v0.3.187: Chats are included in backups, with their own tab and a chat history box.
- v0.3.187: A chat in Trash is not treated as a clash when importing.
- v0.3.187: A chat exported without secrets can be imported again.
- v0.3.187: A move that meets an out-of-date record refuses instead of quietly doing nothing.
- v0.3.189: An agent can delete a folder tree it created.
- v0.3.189: A generated image names its type and size.
- v0.3.191: A generated image keeps its location for the rest of the turn.
- v0.3.192: The Drive connection is ready at start-up rather than on first use.
- v0.3.225: Microsoft 365 mailboxes can be connected, and stay connected.
- v0.3.228: Audio transcription no longer invents a prompt you did not ask for.
- v0.3.230: Trash follows the scope in the header, and All projects can be emptied.
- v0.3.201: A file error names your own path, not an internal one.
- v0.3.204: A chat created and never used can be removed.
- v0.3.237: Cloud storage syncing runs under a memory ceiling, so a large sync cannot crowd out everything else.
- v0.3.238: A text file opened from a workspace keeps its accents and symbols.
Speed and reliability
35 fixes- v0.3.177: Dozens of refusals answer with a sentence instead of a bare code.
- v0.3.181: Creating, updating and deleting agents works on a running installation.
- v0.3.183: Start-up repairs its own settings instead of failing.
- v0.3.184: Hundreds of messages that printed a meaningless value now print the real one.
- v0.3.186: One event-loop monitor per installation, however many times it reloads.
- v0.3.189: Health checks answer before anything else at start-up.
- v0.3.189: Team chat attributes every teammate's message, not just the newest.
- v0.3.194: A chat with no messages yet reads as empty rather than as an error.
- v0.3.194: A busy database waits its turn instead of holding everything up.
- v0.3.195: A chat run that failed reports failure.
- v0.3.195: Compaction attributes teammate messages correctly.
- v0.3.196: Every installation tells its clients which build it is running.
- v0.3.198: Long conversations no longer stall the runtime while being read or written.
- v0.3.198: A brief disconnection no longer removes a computer that has reconnected.
- v0.3.198: The reconnecting screen's logo loads while the runtime is down.
- v0.3.198: A teammate's reply reaches someone who already has messages waiting.
- v0.3.219: A settings write-lock left by a stopped installation is reclaimed safely.
- v0.3.220: Work the platform runs on your behalf is not blocked by a permission you never set.
- v0.3.221: A run interrupted by a restart resumes on the next start-up.
- v0.3.226: One faulty account no longer takes the whole channel status down.
- v0.3.228: A background run that timed out says so, instead of reporting success.
- v0.3.228: A cost refusal names its cause, its numbers and whose allowance it is.
- v0.3.230: A failed import names its cause, with the sensitive part hidden.
- v0.3.231: One channel that cannot describe itself no longer takes the whole status summary down.
- v0.3.231: Health files are replaced cleanly, so a slow check cannot corrupt them.
- v0.3.235: Long chat lists are read a page at a time instead of loading everything.
- v0.3.235: Clearing out expired runs uses an index rather than reading every record.
- v0.3.237: Dozens of operations stopped reading the whole chat store to change one thing.
- v0.3.238: A connection that times out while starting says so plainly.
- v0.3.238: A voice note added between turns is no longer treated as an interrupted reply.
- v0.3.242: A background command's completion still reaches the chat after a restart.
- v0.3.242: A chat gets at most six background-command notices an hour, and the rest as one summary.
- v0.3.243: A search probe that times out no longer reports the index as down.
- v0.3.244: The last whole-transcript reads leave the path every turn takes.
- v0.3.245: Clearing out old sessions keeps to its own steady cadence.
Release
Fixes in v0.3.176
201 fixes across 10 categories.
Chat and interface
30 fixes- v0.3.162: The Admin chat panel sits in the correct place instead of sliding over the page.
- v0.3.162: The Background Services tab is no longer hidden beyond the edge of the header.
- v0.3.163: A chat opened from a notification is refused clearly for an unknown agent instead of failing silently.
- v0.3.163: An opening message is used once rather than being queued again on every reload.
- v0.3.163: A pending message settles correctly when its own send creates the conversation.
- v0.3.165: A chat opened before its first reply is readable now recovers instead of appearing broken.
- v0.3.165: A new chat stays open while its transcript is still being written.
- v0.3.165: A queued model change can be taken back before it applies.
- v0.3.165: Routing notices in chat carry a proper title instead of showing a fault code in its place.
- v0.3.165: The delete progress dialog stays still instead of shifting while it works.
- v0.3.165: Mark-all-read controls in the sidebar work reliably.
- v0.3.166: A new chat's record is written when the reply starts rather than when it finishes, so the conversation is never briefly unreadable.
- v0.3.167: The context gauge measures usage against the window of the model that actually produced the reply.
- v0.3.167: Each usage meter counts against its own model rather than dividing one model's figure by another's limit.
- v0.3.167: Every timer and meter now updates on one shared clock, so figures on a page agree with each other.
- v0.3.169: A reply cut short by a restart keeps the part it had already written, instead of arriving empty or blaming your settings.
- v0.3.169: The conversation records that a restart interrupted a reply, and how much was delivered.
- v0.3.169: A second restart while one is already running joins the one in progress rather than leaving the instance stuck.
- v0.3.169: A chat whose very first message failed can now be deleted.
- v0.3.169: A refused delete or rewind explains why instead of appearing to do nothing.
- v0.3.169: A voice note can carry an attachment in the same message.
- v0.3.169: A message that fails to send while offline returns its text and files to the composer instead of being lost.
- v0.3.170: A missing file is described as a missing file rather than as a technical error code.
- v0.3.170: One broken piece of the interface no longer forces every page to reload.
- v0.3.170: A sidebar chat says which project it belongs to instead of silently assuming one.
- v0.3.176: Preview links open the preview pane again.
- v0.3.174: A prompt card can be closed to skip it, and the agent is told it was closed.
- v0.3.174: A delete started from the sidebar can be minimised while it runs.
- v0.3.174: Only one compaction indicator is shown rather than two.
- v0.3.171: A half-answered prompt survives the next message arriving.
Sign-in, accounts and providers
24 fixes- v0.3.163: Your account name is kept when you reconnect a provider.
- v0.3.163: A newly connected provider no longer reports itself as retrying.
- v0.3.163: Stored sign-in details survive interruption more reliably.
- v0.3.163: Sign-in logs no longer discard the evidence needed to explain a failure.
- v0.3.163: Player One tells you as soon as a provider sign-in stops working.
- v0.3.163: Dead provider sign-ins are shown on the dashboard rather than only failing in use.
- v0.3.164: An expired sign-in token is recognised correctly instead of being misread as a different fault.
- v0.3.164: A provider that reports a dead sign-in is treated as needing reconnection rather than retried indefinitely.
- v0.3.164: A brief refresh delay on one account no longer blocks the other accounts on that provider.
- v0.3.165: Player One tries one more refresh before concluding that a sign-in has failed.
- v0.3.165: Signed-out and Reconnect are only shown when there is real evidence for them.
- v0.3.165: One account's cooldown no longer pauses every account on that provider.
- v0.3.165: Provider health stops blanking out while it rechecks.
- v0.3.165: Reconnecting from an account row carries that intent through the whole flow.
- v0.3.165: A failed account states its condition once, in plain readable text.
- v0.3.166: Every usable account is tried before Player One moves to the next model.
- v0.3.166: A sign-in that cannot be refreshed stays disabled until you reconnect it, rather than being retried.
- v0.3.166: An account with a name shows the name instead of a partly hidden email address.
- v0.3.167: An account that is permanently signed out states that as the reason it was not used.
- v0.3.167: A refresh signal is only sent where something is listening for it.
- v0.3.167: A cooldown countdown reads in days and hours rather than an abrupt short form.
- v0.3.167: A countdown shows precise time while an elapsed time stays approximate.
- v0.3.168: A countdown rounds up, so a nine-minute wait says nine minutes.
- v0.3.168: Provider cooldowns refresh without briefly appearing empty.
Long conversations, compaction and memory
7 fixes- v0.3.162: A failed full search sweep automatically retries after a safe delay.
- v0.3.172: A long conversation is measured against the model actually handling it.
- v0.3.172: The unchanging part of a conversation is measured in the right model's units.
- v0.3.172: A request's history is the conversation itself rather than a summarised view of it.
- v0.3.172: Text shortened for an agent now says that it was shortened, and where the rest is.
- v0.3.172: Per-conversation job detail is kept out of the cached instructions, so the saving holds.
- v0.3.172: A component version that cannot be read keeps its last known value instead of disappearing.
Jobs, coordinators and scheduled work
24 fixes- v0.3.162: A task can complete successfully after deliberately deleting its own temporary output.
- v0.3.162: Automatic job recovery pauses correctly while the main turn queue is under pressure.
- v0.3.162: Scheduled turns are counted correctly and release their capacity after an in-process restart.
- v0.3.163: Cancelling a job reports that cancellation has started rather than claiming it has already finished.
- v0.3.165: A task claim is released when its worker fails terminally, so the work is not left stranded.
- v0.3.165: New workers wait for provider capacity before starting rather than failing immediately.
- v0.3.169: A scheduled task that delivers to a web chat no longer fails silently with no message and no error.
- v0.3.169: A scheduled task with no conversation attached is refused when it is created, rather than delivering somewhere unrelated.
- v0.3.169: A job whose coordinator is stopped by a restart recovers on its own instead of appearing to still be running.
- v0.3.170: A cancellation that was explicitly requested is no longer reported as a failure.
- v0.3.170: A retired session cannot reclaim work it no longer owns.
- v0.3.172: Cancelling a job can no longer leave it half-cancelled.
- v0.3.172: Deleting a job or a run removes its worker records too, instead of leaving them behind.
- v0.3.172: Deleting a job says what happened to its files.
- v0.3.172: A job's files are settled the same way whichever route deleted it.
- v0.3.172: The limit on resuming a job is yours to reset rather than something the job decides.
- v0.3.172: A worker that is temporarily throttled is no longer treated as dead, and a capped resume counts once.
- v0.3.172: A reused worker that fails releases the task it was holding.
- v0.3.172: An attempt that only ran into a throttle is returned to the budget rather than being spent.
- v0.3.172: Job status answers with a summary, with the full detail available on request.
- v0.3.172: Job status keeps the whole list of workspaces instead of collapsing it to one.
- v0.3.173: A scheduled automation with no accepted build is disabled until it is compiled, and says so.
- v0.3.173: A repaired schedule keeps the state it already had.
- v0.3.173: A schedule is registered correctly whatever shape it was written in.
Models, images, PDFs and voice
26 fixes- v0.3.161: Finalised Kimi tool arguments are no longer cleared by the adapter.
- v0.3.162: Player One no longer sends a disabled thinking setting to Kimi models that always use thinking.
- v0.3.162: PDF processing sends subscription credentials using the correct OAuth method.
- v0.3.162: Media tools create the storage locations they advertise and only name locations that exist.
- v0.3.162: Streaming markers no longer become mixed into the displayed model name.
- v0.3.165: A temporary fallback is no longer mistaken for the model you chose yourself.
- v0.3.165: A requested model is only set aside when there is genuine matching evidence for doing so.
- v0.3.165: Player One explains why a requested model did not serve a reply.
- v0.3.165: Choosing to ignore a model reads as your own decision rather than as a failure.
- v0.3.165: A switched-off provider is refused consistently everywhere a model could be selected.
- v0.3.165: Your pinned model is kept when a provider is switched off.
- v0.3.165: The routing chain warning counts exactly what the rows show.
- v0.3.165: The model list keeps its scrollbar clear of the ignore controls.
- v0.3.167: A routing notice explains what was tried and why, instead of showing an internal code.
- v0.3.167: Context overflow is described in ordinary words, with the raw detail kept aside.
- v0.3.167: A fallback notice names what went wrong and the chain it walked.
- v0.3.167: A displaced reply names where it landed as well as what went wrong.
- v0.3.167: Error notices carry a suggested remedy.
- v0.3.169: The developer API reports the model that actually answered, rather than the one that was requested.
- v0.3.169: Asking for a model that cannot be served returns a clear not-found response instead of a server error.
- v0.3.169: A Claude Max subscription is recognised as a subscription rather than treated as an API key.
- v0.3.169: A failed attempt's partial output no longer sits beside the retry's real answer.
- v0.3.170: Generated images are saved into the project you are working in.
- v0.3.174: A temporary provider throttle no longer tells you that you have reached your usage limit.
- v0.3.174: An unexplained provider failure stops blaming your configuration.
- v0.3.174: A connection that drops mid-reply falls back to a second route and reports the provider's own error.
Browser work
10 fixes- v0.3.161: A page that blocks every browser connection can be detected and recovered.
- v0.3.161: When a managed browser dies, Player One explains why and why a replacement was launched.
- v0.3.162: Each chat’s browser tabs remain isolated from unrelated chats.
- v0.3.162: Sub-agents can work inside the browser tabs owned by their parent chat.
- v0.3.162: Player One no longer closes a browser page that the current chat only borrowed.
- v0.3.162: Intentional browser-tab takeover between chats works without confusing ownership.
- v0.3.163: A claimed tab names the chat that lost it rather than one of its helpers.
- v0.3.176: Text can be copied out of the embedded browser while watching it, with an explanation if the clipboard refuses.
- v0.3.173: A closed tab says why it is gone rather than waiting on a page that will not answer.
- v0.3.173: A managed browser profile is cleaned up on the same schedule as the rest.
Email, messaging and connected services
14 fixes- v0.3.161: A cloud-sync process that has already been reaped is reported as unknown rather than falsely failed.
- v0.3.162: Dropbox Business can browse team folders as well as the individual member folder.
- v0.3.162: WhatsApp media failures preserve the real decoding error, improving recovery and diagnosis.
- v0.3.163: An explicit Telegram pin is respected for a direct message.
- v0.3.163: Direct-message senders are kept properly separate from one another.
- v0.3.163: A chat keeps a durable link to the channel it replies through, in both directions.
- v0.3.163: Cloud sync history is kept to a sensible size without losing the baselines it needs.
- v0.3.165: Email accepts vault-backed account credentials and degrades one account at a time.
- v0.3.165: Email states a clear reason when an account is not configured.
- v0.3.169: Two-way folder sync propagates deletions again, instead of quietly growing into a copy of both sides.
- v0.3.169: A safety guard that stops a sync when most files have vanished from one side is switched on, protecting a folder from being emptied by a stale baseline.
- v0.3.169: A folder that has drifted into a growing superset now says so rather than appearing healthy.
- v0.3.169: A pair that has already been set up is recognised as such, so every pass no longer runs as a first pass.
- v0.3.170: An instance-wide connection can be used from any project.
Projects, workspaces and search
18 fixes- v0.3.163: Workspace search genuinely matches on file name as well as content.
- v0.3.163: An agent's chats stay its own, whichever project they were filed in.
- v0.3.163: Project and agent are independent filters, so choosing one no longer disturbs the other.
- v0.3.163: Chat search searches the project rather than a guessed agent.
- v0.3.163: Changing search scope no longer leaves the previous tab's results on screen.
- v0.3.163: The search loading placeholder no longer flashes on a fast response.
- v0.3.163: Search scope tabs stay centred, with the agent filter kept to the right.
- v0.3.163: A dropped connection no longer collapses the all-projects view.
- v0.3.163: The all-projects sidebar no longer re-renders repeatedly.
- v0.3.163: Incoming channel messages are filed under the correct project.
- v0.3.163: Player One warns when a channel route would file its chats somewhere unhelpful.
- v0.3.166: The hourly search update stops re-reading every conversation's reports.
- v0.3.167: Removing an out-of-date search entry is reported rather than done silently.
- v0.3.170: Files are written to the project you are working in, not the agent's own workspace.
- v0.3.170: Commands run in the project's workspace when you are working inside a project.
- v0.3.170: A write that lands outside the current project says so, and names where it went.
- v0.3.170: Search names the projects a shared knowledge source is connected to.
- v0.3.172: Everything an agent creates lands in the project you are working in.
Admin, configuration and setup
28 fixes- v0.3.163: Settings save what you actually chose rather than the underlying defaults.
- v0.3.163: A brand-new instance can write its configuration for the first time.
- v0.3.163: Agent changes can be saved on a fresh instance.
- v0.3.163: Configuration writes are serialised, so two simultaneous changes cannot overwrite each other.
- v0.3.163: The sign-in token is no longer lost during start-up.
- v0.3.163: A saved credential reference is never replaced by its hidden display form.
- v0.3.163: A successful save no longer reports itself as partially written.
- v0.3.163: Editing a channel stays an edit rather than triggering a restart.
- v0.3.163: Configuration backups are kept by age rather than by number of writes.
- v0.3.163: Player One tells you when a configured credential cannot start.
- v0.3.163: A stranded services record is recovered automatically.
- v0.3.163: Creating a sandboxed agent works again.
- v0.3.163: A sandboxed agent's limits cannot be widened by configuration.
- v0.3.163: A sandboxed agent is no longer offered context it cannot reach.
- v0.3.163: A sandboxed agent's available tools are now an explicit list rather than inherited automatically, so a tool added for general convenience no longer reaches customer-facing agents.
- v0.3.163: A sandboxed agent can use every tool it is permitted, rather than only those its stored settings happened to name.
- v0.3.163: A tool that resolves access from the caller is no longer available to sandboxed agents, so operator automations cannot be exposed publicly.
- v0.3.163: The sandboxed marker on an agent stops disappearing between health updates.
- v0.3.163: Health checks no longer report a working sandboxed agent as broken when run without elevated permissions.
- v0.3.163: Deleting an agent clears its links through one consistent rule.
- v0.3.165: Queued commands show live progress and can be withdrawn.
- v0.3.169: Provider switches record who changed them and when.
- v0.3.169: An out-of-date removal record can no longer delete a credential you have just reconnected.
- v0.3.169: A tool that is excluded by design says so, rather than looking like a permission you could grant.
- v0.3.170: Setup steps follow one consistent route rather than several.
- v0.3.170: A refused vault request explains the reason instead of returning a single word.
- v0.3.170: Access approvals appear in the chat that asked for them.
- v0.3.170: An approval names the project it applies to.
Resources, stability and security
20 fixes- v0.3.162: Disk-pressure queue holds are applied where disk state is actually decided.
- v0.3.162: Internal scheduled callbacks retain the operator authority they were given.
- v0.3.162: Vault references are rejected from unsafe environment-variable injection paths.
- v0.3.162: API keys that cannot incur spend explain that clearly instead of showing an empty figure.
- v0.3.163: Player One recovers on its own if its main loop becomes blocked.
- v0.3.163: A command stopped for using too much memory now explains how to recover.
- v0.3.163: Log searches no longer loop over the same section indefinitely.
- v0.3.165: The sign-in page keeps checking calmly while an instance wakes, and explains a longer wait.
- v0.3.167: Background work no longer disturbs the chat list or its ordering.
- v0.3.167: Internal routing detail is no longer broadcast to the operator interface.
- v0.3.167: The agent roster loads without its full configuration detail until that is asked for.
- v0.3.166: A refused agent-to-agent message explains how to send it correctly.
- v0.3.169: A restart is attributed accurately, using the system's own record rather than guesswork.
- v0.3.169: Indexing no longer rebuilds an entire workspace to notice one changed file.
- v0.3.169: A settings store that cannot be read says so instead of quietly serving defaults.
- v0.3.170: A queued message frees its place as soon as it starts, rather than at the end of the reply.
- v0.3.172: A tool called through the API is checked for its required fields before it runs.
- v0.3.171: A workspace grant can never reach the credential store.
- v0.3.171: A damaged credential store costs only the vault rather than the whole instance, and permission problems are reported loudly.
- v0.3.173: Setup works on the newest Ubuntu release.
Release
Fixes in v0.3.160
103 fixes across 7 categories.
Chat and interface
18 fixes- v0.3.147: New chats no longer remain frozen briefly after being created.
- v0.3.147: Returning to a recent chat includes messages that arrived while you were elsewhere.
- v0.3.147: Malformed older conversation records are isolated instead of breaking the complete chat history.
- v0.3.147: Older child conversations are found correctly when checking agent and worker relationships.
- v0.3.147: Job tabs no longer jump during navigation and now have stable routes.
- v0.3.147: The side panel opens the current project’s default workspace instead of an unrelated workspace.
- v0.3.147: System version information remains visible and reports the actual gateway version.
- v0.3.147: Invalid form definitions now produce a correctable message instead of failing unclearly.
- v0.3.148: Aborted responses no longer leave orphaned tool results that can damage the conversation.
- v0.3.148: Reserved Agents pages no longer scope the sidebar to a non-existent agent.
- v0.3.149: Conversation and agent-history disagreements are reported instead of silently swallowed.
- v0.3.150: A renamed chat immediately shows its new title inside folders.
- v0.3.150: The Admin mark keeps the correct spacing when its menu scrolls.
- v0.3.152: Chats open inside the project they actually belong to.
- v0.3.152: Chats rejected by Anthropic because of a malformed error result can recover instead of remaining unusable.
- v0.3.152: Search excerpts stop at whole words rather than cutting words in half.
- v0.3.152: Generated image replies show the created filename clearly.
- v0.3.152: Memory counts no longer bunch together on narrow screens.
Long conversations, compaction and memory
17 fixes- v0.3.148: Compaction retries temporary provider failures instead of abandoning the conversation.
- v0.3.148: OpenAI and Codex context-limit messages are recognised correctly.
- v0.3.148: Oversized conversation sections can be divided even when the first error is unclear.
- v0.3.148: Very large thinking blocks are capped before they inflate the compaction request.
- v0.3.148: Compaction keeps its start and end boundaries within safe limits.
- v0.3.148: Citation ranges include the final section of the conversation.
- v0.3.148: Divided sections omit uncertain citations rather than inventing them.
- v0.3.149: Compaction progress remains visible if the page is opened or refreshed while compaction is running.
- v0.3.149: Different parts of the platform now share one bounded view of active compaction work.
- v0.3.149: Forced text splitting no longer tears code-fence markers.
- v0.3.152: Modernising an older Memory Space preserves its Fold index correctly.
- v0.3.152: Memory migrations now state accurately when reindexing is required.
- v0.3.152: Cross-project memory refusals name both involved projects, making the issue understandable.
- v0.3.154: Dead workspace search bindings are repaired automatically.
- v0.3.154: Older skipped-file markers are healed without stalling partway through the backlog.
- v0.3.158: Workspace file contents are included properly in search.
- v0.3.158: Search reindexing remains visible from its start through any failure.
Jobs, coordinators and scheduled work
15 fixes- v0.3.151: A coordinator with an approved task dispatches it instead of returning to planning.
- v0.3.151: A coordinator whose run temporarily has no tasks can wake and continue correctly.
- v0.3.151: Scheduled reports are delivered even when the underlying run has nothing else to say.
- v0.3.151: Quiet-run notices include the schedule name, so the source is clear.
- v0.3.152: Cancelling a scheduled task stops the run that schedule started.
- v0.3.152: Cancelling a job stops its coordinator as well as its workers.
- v0.3.152: Cancelled jobs can resume with their existing workers.
- v0.3.152: Finished runs remain valid places for a job to resume.
- v0.3.152: Resuming a job no longer creates a coordinator inside a run that does not exist.
- v0.3.152: The coordinator resumes immediately and steers existing workers instead of unnecessarily replacing them.
- v0.3.152: Cancellation reports that a stop was requested rather than falsely claiming everything was killed instantly.
- v0.3.152: Provider-capacity waits are visible in job status.
- v0.3.152: Provider outages wait safely instead of being hammered with repeated requests.
- v0.3.152: Job completion checks the approved plan rather than relying only on the task board.
- v0.3.152: A delegated helper is stood down before another agent takes its work back, avoiding duplicate output.
Models, images, PDFs and voice
10 fixes- v0.3.147: Disconnected providers are removed from active model-routing chains.
- v0.3.147: A temporarily quarantined model is moved down the route rather than disappearing completely.
- v0.3.147: Live calls no longer expose the wrong text-to-speech tool.
- v0.3.148: An unsubscribed provider is no longer incorrectly described as absent from the catalogue.
- v0.3.148: Image edits no longer fail because of an inferred resolution the user never selected.
- v0.3.148: Camera captures are consistently treated as images.
- v0.3.149: Corrupt images produce an accurate error instead of being blamed on file size.
- v0.3.151: Model fallback checks whether the replacement supports the required input type.
- v0.3.151: A model fallback no longer silently removes the agent’s tools.
- v0.3.152-v0.3.153: Model-switch dividers identify when an agent changed the model itself.
Email, messaging and connected services
18 fixes- v0.3.147: Failed channel-recovery loops terminate instead of retrying forever.
- v0.3.147: Discord has stronger protection against a channel-library crash.
- v0.3.149: A disabled Telegram account keeps its update position, preventing old messages being replayed later.
- v0.3.149: Google Drive connections can reach the upload path correctly.
- v0.3.149: Cloud sync explains when there was nothing to upload.
- v0.3.149: A zero cloud-storage measurement is shown as unmeasured rather than empty.
- v0.3.149: Cloud sync explains where files are missing from.
- v0.3.149: Writes that would be hidden by a synced cloud folder are refused rather than silently lost.
- v0.3.149: Each workspace is watched once rather than being registered repeatedly.
- v0.3.149: Public webhook paths now match the paths actually served by the gateway.
- v0.3.149: Agents can tell when inbound webhooks are disabled or unavailable.
- v0.3.151: Matrix resolves its password correctly when a token is used for authentication.
- v0.3.151: Email-triggered conversations attach to the correspondent rather than the operator’s mailbox.
- v0.3.151: Email activity reports the correspondent as the conversation source.
- v0.3.151: A saved Vault secret remains stored until the user explicitly removes it.
- v0.3.155: The email channel resolves its platform dependencies correctly after installation changes.
- v0.3.160: Cloud sync excludes Git metadata instead of copying repository internals.
- v0.3.160: Webhook names are decoded safely and invalid targets are refused clearly.
Projects, workspaces and search
11 fixes- v0.3.151: Project search ranks results inside the authorised project set.
- v0.3.151: Projects reject membership entries for entities that do not exist.
- v0.3.152: New workspaces no longer bind themselves to unreachable projects.
- v0.3.152: Workspace refusals explain which check failed instead of returning one generic error.
- v0.3.157: Moving resources out of the Default Project preserves their existing access.
- v0.3.157: Project membership validation correctly includes scheduled tasks and flows.
- v0.3.157: The orphan-repair sweep can claim all resource types it needs to fix.
- v0.3.157: Project membership rejects agent identifiers that do not name a real agent.
- v0.3.158: Bulk pause and resume only act on Fold sources that can actually change state.
- v0.3.158: Unmanaged-resource reports remain small enough to read and can be searched by name or identifier.
- v0.3.160: The mobile assistance chat now mounts and appears correctly on phones.
Resources, stability and security
14 fixes- v0.3.147: Per-run memory is released correctly, addressing long-uptime out-of-memory restarts.
- v0.3.147: Listing conversations no longer starves other gateway work.
- v0.3.147: Routine health checks stop reading from disk after warm-up.
- v0.3.149: Avatar paths can no longer escape through a symbolic link.
- v0.3.149: Browser security policy and lock-failure containment were tightened.
- v0.3.152: The Resources page distinguishes unreadable folders from folders that have not been scanned.
- v0.3.152: Empty resource rows are removed instead of showing misleading entries.
- v0.3.152: Disk usage measures the whole disk accurately.
- v0.3.152: Installed tools are no longer mistaken for disposable cache during storage cleanup.
- v0.3.160: Old temporary package installations are dated correctly before cleanup.
- v0.3.160: Package cleanup fails safely if it cannot confirm what should be removed.
- v0.3.160: The storage bar has enough room and explains what the remaining space contains.
- v0.3.160: Gateway close and boot operations no longer wedge a serving instance.
- v0.3.160-v0.3.162: Failed restarts exit cleanly instead of leaving a broken process occupying the port.