
Track work in five steps
Open the board
Cmd+2 (Ctrl+2 on Windows and Linux). The board opens on Kanban the first time, and on the view you picked after that. The team filter beside the Board title shows All teams or one team.Read a card
Act on Needs you
Open a task
Find a task
/ while the board has focus, or press Cmd+K (Ctrl+K). Type part of a title, an agent’s name or words from the details. Enter shows the card on the board, and Shift+Enter opens its drawer.The header and toolbar
The header holds the Board title, the team filter, and New task (pressC). When something changed while you were away, a chip beside the team filter says what; see Since you last looked under Views and the calendar.
The toolbar under it holds these controls, left to right:
Columns and statuses
Your requests and their sub-tasks
When a team’s lead hands out work for something you asked in chat, the board shows your request as one card, with the helpers’ tasks under it. The card says how far the work has got, like 2 of 3 done, and its cost is what its work has cost so far. Click 3 sub-tasks to fold or open the list, and click a row to open that sub-task. A finished request starts folded. The card moves by itself: it sits in In progress until everything under it has closed and the lead has written its summary, then moves to Done, or to Cancelled if none of its work got done. You cannot drag it. To end it early, open it and set its status to Done or Cancelled: the sub-tasks still open are cancelled too, and any run on them stops. The board asks first. To add to a request, type in its reply box. Your words go to the agent looking after it, which can hand out more under it. To say more on a finished one, Reopen it with your words: it goes back to In progress, its lead reads them, and it closes again after the lead’s reply unless that hands out more. A new message in the chat is a new request. A sub-task that needs you also shows in Needs you, with For CLB-12 under its title to say which request it belongs to. If you filter the board so a request is hidden but one of its sub-tasks is not, the sub-task shows on its own with the same label.Read a card
Cards are neutral: color appears only in the status marks, and amber marks what needs you. A card shows the task’s title, on up to two lines, and one row of details:- Id: the task’s readable id, like
MAR-12. See Task ids. - Agent: the face of the agent that has the task. In To do, a dashed circle means no agent is bound to it.
- Priority: two small up chevrons for an Urgent task, one for High. A Normal or Low task shows nothing here. Point at the mark to read it.
- Status mark: Verifying while the work is checked, and Waiting to check while its check waits for teammates to finish in the team folder. After a check, an amber cross means it failed, an amber dashed circle means it passed with debt (open issues noted), and a green check means it passed (until the card reaches Done). A card with no mark shows its cost here.
- Cost: what all of the task’s runs cost, once a run has recorded a cost. A figure that starts with
~is an estimate: the runtime reported no charge, so Clawboo priced the run itself. A cost under a cent reads<$0.01, and a task whose runs all ran on a model your own Ollama serves readsfree. When a status mark takes the cost’s place, point at the card to see the cost. Which runtimes report real cost is in Budgets and spend. - Paused: a small pause mark when a pause covers the task. Point at it to read what is paused and who paused it. The face of a paused agent carries the same mark wherever it shows.
- Subtask: a small arrow when another task handed this one out and the card has no mark or cost to show. Otherwise the mark’s or the cost’s tooltip says it is a subtask.
- Labels: up to two of the task’s labels, then a count of the rest. See Labels.
- Time: on Kanban, how long ago the task was added or finished, such as “4m ago”, “Yesterday” or “3d ago”. A calendar view shows the clock time, and the Earlier row the date. A running task shows how long its current run has gone, after a small live dot. Point at the time for the full date.
Things that need you
Needs you collects everything that waits on a person, for the teams in the filter. It holds two kinds of item:- Approvals: an agent wants to run a command, use a risky tool, or make a risky handoff. Answer them on their card. See Approvals.
- Tasks that will not move by themselves. Each card says why in amber, with the reason the run recorded when there is one, and shows the agent.
Get an alert outside the tab
Clawboo can send a browser notification when something new needs you while its tab is in the background or its window is not in front. Once the board has a task on it, the header offers Get an alert when something needs you, even with this tab in the background. Click Turn on, then allow notifications when your browser asks. Not now puts the offer away. Clawboo asks your browser only when you click Turn on, never by itself.See it all in the Inbox
Inbox in the sidebar is the board from your side: what wants you, whichever view you are on. Its red badge counts what is unread and what needs you, each thing once, and goes up by itself as things change. While the sidebar is collapsed, or closed on a phone, a red dot on its button says there is something to see.When a task runs again by itself
If a run crashes or stops responding and its task goes back to To do, Clawboo starts it again by itself, through the task runner for a task the runner ran. A run whose sign-in or billing was refused is the exception: it goes to Needs you at once. After three unsuccessful runs in a row it stops trying, and the task waits in Needs you as Failed 3× or Timed out 3×. Crashes and failures are counted apart, and it stops at 3 in a row of either. A task you stopped never restarts by itself. A budget cap or the circuit breaker stopping a run is not your Stop, and the card says which it was. A task that was in review when Clawboo restarted does not run again by itself either: it waits in Needs you with a comment saying its run stopped. Retry always runs a task again, whatever its count. In team chat a run’s ending decides what comes next, so a retry happens only when it can help:- The provider rate limited the run, was overloaded, or could not be reached. The task goes back to To do and says so, then runs again after 30 seconds and after 60 seconds of time the machine is awake, with a little added. It holds no run slot and no budget while it waits, a restart picks up where it was, and Stop cancels the wait. After two retries it waits in Needs you as Rate limited or Provider unavailable, and whoever handed it out is told. A task on the task runner waits at least as long before each automatic rerun. A comment from you starts it at once. A task that has already failed 3 times in a row does not retry by itself, so a Retry that hits the same limit goes straight back to Needs you.
- The agent replied with nothing and has nothing else to show. It gets one more turn, asked to say what it did. A second empty reply waits in Needs you as No result. A reply of only a control token such as
__skipped__counts as empty. An empty reply reaches Done in one case: the run changed the team’s folder and the team’s work check passed it. If a teammate was working in the folder at the same time, the change may not be this agent’s, so it is asked first. - The agent ran out of room. It gets one more run to continue, and what it wrote so far stays on the task. A second time waits in Needs you as Out of room. Each kind of extra turn happens once until something new starts the task, so the two cannot hand each other more.
- A hard budget cap stopped it. It waits in Needs you as Budget cap. Raise the cap, then Retry. When the rest of a budget is only set aside by runs already going, the task waits in To do instead, and starts by itself once they end. If what they spent fills the cap, it comes to Needs you as Budget cap, and whoever handed it out is told.
Give an agent a task
Open New task
C. The + in the To do column’s title opens it too.Say what should get done
Check the agent
Add details if you need to
Send it
- It matches the words of your sentence against each agent’s name and the role in its SOUL.md (see Agents). A word in the name counts more than one in the role, and forms of a word count as one, so “drafting” matches “drafts”. OpenClaw agents, and agents whose SOUL.md states no role, match by name only.
- When nothing matches, it suggests the agent you last gave a task to in this browser, then the team lead, then the team’s first agent.
- The list holds the agents of the team in the board’s team filter. With All teams, it holds every team’s agents, each with its team’s name.
- Only that agent works on it. The agent is asked to do the work itself rather than hand it to a teammate. It is asked to make reasonable assumptions and note them instead of asking you questions, though you can still steer it with a reply.
- Its result goes on the task’s card, on the board and in the team’s General chat. It is not passed to the team leader.
- If the run fails, the card moves to Needs you with the reason.
Steer a running task
Reply on a task to talk to the agent it belongs to. While the task runs, the agent reads your reply during the run where its runtime allows it. When the task waits on you, your reply starts the agent on it again, so answering a blocked task needs no click on Retry.Open the task
Check what sending does
Write and send
Shift+Enter adds a new line).Pause work
A pause holds back one agent, one team, or a task with the tasks under it, until you resume it. It lasts through a restart, and it has no time limit.- Runs stop within seconds. A task whose run stopped goes back to To do and keeps its agent. The task’s comments say why. It does not show in Needs you. A routine’s run that stops does not count as having run: it runs again on resume, as a new task, and its stopped card is set aside.
- Nothing it covers starts. No task starts, no Routine fires and no message wakes it. A run already waiting for a free run slot stays waiting. Retry says the task is paused.
- Messages get an answer. A message to a paused agent or team stays in the chat, with a line saying it did not start. The agent does not get it, so send it again after you resume. A reply on a paused task is saved, and the note under it says so. So is your answer to an agent’s question, and a note on a task you reopen. The agent starts on them after the resume, on a task in Needs you too, and the task drawer says what it waits for. A comment you write on a task whose agent asked you a question counts the same as with no pause: it starts the task, and the question closes unanswered.
- Tasks keep their place. Each task stays in its status, and you can still assign tasks to a paused agent. They wait in To do. A task an agent asked you about still shows its question in Needs you, and you can answer it during the pause. A task you took yourself stays yours: it reads as neither paused nor started by a resume. A task waiting on work a pause holds keeps waiting and reads as needing nobody, and its agent takes its turn after the resume, with every result that came back.
- A 1:1 chat with an OpenClaw agent runs in the OpenClaw Gateway, so it still works. The chat’s banner says so.
- A team’s work check that is already running finishes its command. It starts no reviewer under the pause, and the result says the review did not run. If the check fails, the task waits in To do with what it found.
- Pausing a team does not pause its agents’ work for another team, and it does not change their tasks’ priority.
The task drawer
Click any card to open its drawer on the right. It reads like a document: the task’s title, with its status, agent and team under it. Press Escape or click outside it to close it. To step through the card’s column without closing the drawer, pressJ for the next task and K for the previous one, or use the up and down arrow buttons at the top of the drawer. The drawer shows where you are, for example “2 of 5”. The keys do nothing while you type in a field or have a list open.
Each section folds from its heading, and a section with nothing in it is not shown. Top to bottom:
free for a task that ran only on a local model.
An image in the agent’s report or its comments shows as a link named after its alt text, such as Image: sales chart. It opens in a new tab only when you click it. Clawboo never loads it by itself, so an agent cannot send data out in an image’s address without your click. An address that is not a web or same-site one is only named, not linked.

Change a task’s status
Use Status in the drawer’s Properties, or drag the card to another column. Both follow the same rules:- Only moves the board allows are offered. While you drag, columns that will not take the card are dimmed, and so is Needs you. Done and Cancelled cards cannot be moved.
- In progress is set only by an agent’s run, so you cannot move a card there by hand.
- Moving a card to Done or Cancelled from a day other than today asks which day it finished: Today, the day it sits on, or Another day.
- Cancelling always asks first, because it cannot be undone. Later plan steps that were waiting on the task are cancelled with it.
- If an agent is still working on the task, the prompt says so. Moving the card to Done, Cancelled or To do stops that run. When an agent handed the task out, that agent is told a person decided it.
- Your move stands. After you move a card, its agent cannot change the task’s status, block it or unblock it until a new run starts on the task. The agent is told the status is yours to set and not to try again. A run that starts after your move works the task as usual.
- A card moved back to To do goes back to its agent’s queue, and a run that was going starts over from the beginning. A card with no agent waits in Needs you as Unassigned.
- If the task’s stored check result cannot pass, moving it to Done asks you to confirm with Complete anyway. The override is recorded in the audit log.
Reopen a finished task
A Done or Cancelled task’s drawer has Reopen. Say what should change, if you like, and it goes back to To do, without its finish date. Its agent starts again, in a fresh session, with your words. In team chat it also has the task’s comments in front of it. A run on the task runner gets your words from its mailbox and no comments. Reopening a task assigned to you gives it back to you, and a task with no agent waits in Needs you for you to assign one. Plan steps cancelled along with it stay cancelled. A step whose earlier step stays cancelled cannot be reopened, because nothing would ever start it: Reopen says so, and you reopen the earlier step first. Agents have no tool for it, and the route takes only a request from the board’s own page, which on a local install is a speed bump.Set a task’s priority
A task has one of four priorities: Urgent, High, Normal (every task starts there) and Low. Only you can set Urgent: an agent creating a task stops at High. Change it with Priority in the drawer’s Properties. When an agent has several tasks waiting, it starts one that already ran and came back first, then the highest priority, then the oldest. A task that keeps waiting reads one priority higher every 4 hours, up to High, so a Low task gets through newer work in the end. Only a task you marked Urgent reads as Urgent, so it always goes first. Priority also counts when every run slot is taken: an Urgent task goes ahead of runs that queued before it. An agent that creates a task through its Tasks tools can set its priority too, as a number from 1 (High) down to -1 (Low); any other number is held to that range, so no agent’s task can jump ahead of an urgent one. The same limit applies to anything that sets a priority without going through this page, such as a routine created through the API.Task ids
Every task of a team has a readable id, on its card, on its row in Needs you and the Inbox, at the top of its drawer and on its card in team chat. Tell an agent “look atMAR-12” and it can find it. Find matches an id too.
The letters come from the team’s name. To use others, open the team’s settings from its chat header and change Task ids. You can do that until the team’s first task exists. After that the letters are fixed, and renaming the team changes nothing. Two teams cannot share letters: Clawboo offers the next free set, like MA2, and refuses a set another team uses.
A task with no team has no readable id. Tasks from before this feature got their ids in the order they were made, the first time Clawboo started after the update.
Labels
Labels are a team’s own words for sorting its board. Set them with Labels in the drawer’s Properties: tick the ones you want, or type a new name. To rename or remove a label for the whole team, open the team’s settings and go to Labels. Removing one takes it off every task. Filter lists the labels on the tasks in view. Under All teams, two teams’ labels with the same name are one entry. Agents do not set labels.Review what an agent built
The Files section shows what a task’s runs made, in one of two places:- A git worktree: a separate checkout of a repository, made for that task. That happens for runs started through the API with a repository path, such as an agent routine created through the API.
- The task’s own folder,
teams/<team id>/tasks/<task id>in the work folder (~/clawboo-workby default), for a task a Claude Code, Codex, Hermes or Clawboo Native agent runs without a worktree: work handed out in team chat, tasks you give from the board, and routines you create in the dashboard. The run is told to save the files it makes for the task there, and that the team’s shared folder is two levels up (../..).
- the Branch (
clawboo/task-<id>), the Worktree folder, and Changes: how many files changed, and the lines added and removed; - under System-of-record files, the files Clawboo keeps in the worktree to track the task, such as
TASK.mdandVERIFICATION.md, each folded; - the full Diff against where the branch started;
- a live view of the worktree’s files, with change badges and what the agent touched last, when the task has an agent;
- a Preview of the page the agent built.
index.html in the worktree or the task’s folder once the agent writes one, served as static files. Scripts do not run in the preview, so an app that needs JavaScript shows only its static shell. Open shows the page in a new tab, where its scripts stay off too. The preview checks for changes every 10 seconds while it is on screen.
Clawboo does not merge the branch. Review it and merge it yourself.
When work is checked
Clawboo checks an agent’s work in two cases: a run in its own git worktree that changed files, and team chat work (a hand-off, or a task you give from the board) that changed the team’s folder on a team with a work check, by an agent that works in that folder, so not an OpenClaw agent (see which team chat work is checked). The check runs the verify command and, for a larger or risky change, or a task handed down from another task, a separate review by a model. Clawboo knows your project’s test command only if you set the team’s work check. Without one, a worktree’s check fails until theVERIFY_CMD line in its init.sh names one, and team chat work is not checked. Agent routines you create in the dashboard are not checked. A task with no check result reaches Done normally.
While a check runs, the card shows Verifying. In team chat a check waits while a teammate is still working in the team folder, or another task’s check is running, and the card shows Waiting to check until then, for 15 minutes at most. A comment on the task says who it is waiting for. A pass lands in Done, and a failed check sends the work back to the agent for another attempt. When the attempts run out, or the work passes with debt, the task goes to Needs you. It shows Needs review if its result cannot pass, or Blocked if it passed with debt. Retry runs the task again and clears the old result. See Verification.
Decide on a plan
To decide before an agent splits up work, open the team’s settings and turn on Ask me before splitting work, under Plans. It is off at first. It covers a plan, a list of tasks an agent proposes, and two or more<delegate> tags in one reply. A hand-off an agent makes by tool call, or one tag alone, still goes straight through.
With it on, the work shows in Needs you as Waiting for your answer, with an alert if you have them on, and lists each step and who it goes to:
- Read the steps. Each box holds the whole of what the agent will be told. Open the card and pick Plan for the document, if the agent wrote one.
- Change a step’s words, remove a step, or add one with Add a task. What is in the box when you accept is what runs, and nothing you took out reaches the agent. A step you add goes to the agent that proposed the plan.
- Press Accept. The steps go on the board in that order, and the first starts. Tasks from several
<delegate>tags all start at once. Or press Turn down, and nothing is created.
Routine runs
In Today, Week and Month, the upcoming runs of your routines appear in To do on the days they will run. Kanban has no days, so it shows no runs. A run is not a task yet, so it is drawn as a dashed strip with the routine’s name and time, not as a card. The next run on screen is drawn solid. A routine that runs several times a day shows once that day, with the number of runs. Click a run to see the routine’s schedule in plain words, where it goes, its next runs, and what happens when it runs. Run now starts it at once, and Open routine opens it to edit, pause or delete it. When an agent routine runs, it becomes a task card on that day. A team routine posts to the team’s General chat, and if its turn hands out work, that run shows as one request card with the work under it. Only routines that will run show runs. A paused routine, one halted after a failed run, and a one-time routine that already ran have none, and neither do the OpenClaw Gateway’s own scheduled jobs. Routine runs under Display shows or hides the runs, and its note says how many are coming up on screen. Times are in your local time, and when routines run in another time zone (see the time zone), the schedule names that zone.Views and the calendar
Pick a view from the view menu at the right of the toolbar, which names the view you are on, or press its key:- Open work sits on the day it was added, and stays on that row as it moves from Backlog to To do to In progress.
- Finished work sits on the day it finished.
- Open work from before the range on screen is gathered in an Earlier row at the top, so nothing in play is hidden. It opens by itself when it holds running or waiting work.
Reading the calendar
Reading the calendar
- Lanes and cards are neutral. Each column’s title carries its status mark in the column’s color, and Needs you is amber.
- Today’s date sits in a red circle. Weekends and the Earlier row are shaded.
- Pick a date in the day column to light its whole row. Pick it again, or press Escape, to put it out.
- A day with nothing on it folds into a thin line, and empty days in a row share one line. Click the line to open those days. Today never folds.
- A cell shows three cards (two in Month view), then +N more, which opens the rest in place. Today and Kanban show every card.
- A card an agent just added or moved fades in, so you can see what changed.
- A move you make shows one message saying what happened. When the card went to a day the board is not showing, the message has a Show button that takes you there.
Moving through time
Moving through time
- In Week and Month, the Today button goes to the current week or month and brings today’s row into view. While today is out of sight, a small arrow on the button points the way.
- The arrows beside it step one week or month back or forward.
- Click the range title, for example “Oct 25 to 31, 2026”, to pick a date. A dot marks each day that holds a task. In Today view, picking a date opens that date’s week.
Keyboard shortcuts
Keyboard shortcuts
Find
Find
- Enter shows the task: the board moves to its day and highlights the card.
- Shift+Enter opens the task’s drawer.
- Enter on a routine opens the routine.
- Escape closes Find.
Since you last looked
Since you last looked
Tasks in team chat
The same tasks appear in the team chat they came from: as cards in the conversation, and in the chat’s Tasks card. A request shows as one card in the conversation, with each sub-task’s own card inside it, and in the Tasks card its sub-tasks sit right under it. General shows the tasks handed out in General, and each room shows only its own. The expand icon on a chat card, and Open on Board in the Tasks card, open the task here with its drawer open. See Team chat. A task handed off from Boo Zero’s own chat belongs to the team’s General, so it shows there as well as in Boo Zero’s chat, and its result goes back to Boo Zero’s chat. The message it came from is a request with its own main task, which closes after Boo Zero’s reply in that chat.Troubleshooting
A task has been In progress for a long time
A task has been In progress for a long time
A task reached Done with no check mark
A task reached Done with no check mark
I cannot move a card to In progress
I cannot move a card to In progress
Cards are missing from the board
Cards are missing from the board
A routine's runs are missing from the calendar
A routine's runs are missing from the calendar
The board says it couldn’t load
The board says it couldn’t load
Under the hood
Under the hood
- The board reads
GET /api/board, every five seconds while the tab is visible. The drawer readsGET /api/board/:taskId,GET /api/board/:taskId/executionsandGET /api/board/:taskId/workspace/detail. Its file view reads theGET /api/board/:taskId/workspace/*file routes for a worktree, andGET /api/board/:taskId/folder/treeand/folder/filefor a task’s own folder. The preview is served fromGET /api/board/:taskId/preview/, out of the worktree or the task’s folder. - Reopen is
POST /api/board/:taskId/reopen, assigning a task to yourself isPOST /api/board/:taskId/assign-person, and answering a question isPOST /api/board/questions/:questionId/answer. Each answers only a request the board’s own page sent. That check is a speed bump on a local install, because an agent running as you can send the same headers. Behind a sign-in proxy inrequiredmode, every request also needs a verified person. See Security. - New task is
POST /api/boardwith anassigneeAgentId, orassignToMefor “You”. It reads each agent’s role fromGET /api/teams/member-roles. Retry, Assign and replies arePOST /api/board/:taskId/retry,/assignand/comments; the reply’s answer carries adeliveryfield that says where it went. A status change isPATCH /api/board/:taskId, and a confirmed Complete anyway sendshumanOverride, which is written to the audit log. - On a team chat run the team engine claims and moves the task itself, so the agent gets no claim or status tools there except
block_task,ask_personandhand_back_taskon its own task. It can also say how the task went withreport_result, which moves nothing. An OpenClaw agent gets none at all, since its tools cannot tell one agent from another. Cancelling a task withupdate_task_statuscancels the plan steps waiting on it, as Dismiss does. - Retry, Assign, and a reply that would start the task again, refuse a task still filed under a deleted chat room (
room_archived) instead of saying it started; restarting Clawboo moves it to General. A task whose team was deleted is refused the same way (no_team), and nothing can run it. - A status change you make records its time on the task (
personMovedAt). Until a run starts after that time, the agent’supdate_task_status,block_taskandunblock_tasktools are refused. See the tasks tools. - The view and the Display choices are stored in this browser’s local storage under
clawboo.board.calendar, and the agent you last gave a task to, per team, underclawboo.board.new-task-agents. Under a path prefix (CLAWBOO_BASE_PATH), each key starts with the prefix, such as/clawboo:clawboo.board.calendar. GET /api/board/:taskId/workspace/detailanswers{ "ok": false, "reason": "no_workspace", "taskFolder": { "path": "...", "onDisk": false, "holdsFiles": false } }for a task with no checkout of its own, and a404only for a task that does not exist.taskFolderis the task’s own folder:onDiskturnstrueonce a run has made it, andholdsFilesonce anything is in it. The drawer shows Files when the folder is on disk and holds files, or while the task is in progress or in review. Its file view reads that folder directly, so an older task’s files show there even when the agent’s Computer tab leaves the task out.- Browser alerts add no polling. They follow the event stream the dashboard already keeps open (
GET /api/obs/stream). When a run ends, a task moves or an approval opens while the tab is in the background, they readGET /api/boardandGET /api/tools/approvals?status=pendingonce, a second after the first event, so a burst of events is one read and one alert. A read that fails is tried again after 5, 20 and 60 seconds, and a shell approval the Gateway delivered is still announced when the approval queue cannot be read. Your choice is stored in this browser’s local storage underclawboo.alerts.needsYou. The desktop app is detected by thewindow.clawbooDesktopbridge it gives the dashboard, and a click on one of its notifications reaches the dashboard through that bridge’sonOpenTarget(see Open the dashboard from a notification). - Full request and response shapes are in the Board API. How the board stores tasks and recovers stuck work is in Board internals.
Next steps
- Team chat: give a team work and follow it beside the conversation.
- Verification: when work is checked, and what to do with a result that cannot pass.
- Worktrees and handoff: where an agent’s file changes go.
- Routines: set up the recurring work that shows on the calendar.
- Approvals: answer the requests that appear in Needs you.
- The board: statuses, dependencies and recovery behind the cards.