What it does
List tasks assigned to an AI agent (your work queue), highest priority first. Call this first each run. By default this is your ready-to-work queue: it excludes done and pending_approval tasks (unless include_done/include_pending_approval), tasks whose plan you've proposed and are awaiting a human on (plan_proposed, unless include_plan_proposed), tasks still blocked by an unfinished dependency (unless include_blocked), and tasks a human or an agent has paused (unless include_paused). A task returning here with plan_approved_at set is one whose plan a human just approved — that's your go-ahead to execute it. When the result is empty, the response includes a hidden breakdown (blocked/done/wont_do/pending_approval/plan_proposed/needs_input/paused/claimed_by_other counts) so you know whether there's no work at all or work that's just filtered out right now — an empty queue where everything is paused is a different situation from an empty board, and this is what tells them apart. Tasks a human marked “won’t do” never come back, with or without include_done. The queue spans every workspace this account belongs to — pass workspace_id (from list_workspaces) to work one board at a time.
Parameters
| Name | Type | Required | Description |
|---|---|---|---|
include_done | boolean | Optional | Include tasks already marked done |
include_pending_approval | boolean | Optional | Include tasks you've already proposed as complete and are awaiting approval |
include_plan_proposed | boolean | Optional | Include tasks whose plan you've proposed and are awaiting a human's approval |
include_blocked | boolean | Optional | Include tasks blocked by an unfinished dependency |
include_needs_input | boolean | Optional | Include tasks parked on a question you haven't answered yet |
include_paused | boolean | Optional | Include paused tasks. They still can't be worked — this is so you can report what you skipped without pulling the whole board |
workspace_id | string | Optional | Only tasks in this workspace. Optional — omitted, the queue spans every workspace this account belongs to, so an account that shares work with other people gets their workspaces' Claude tasks too. Ids come from list_workspaces. |
status | string | Optional | Allowed values not_startedin_progressplan_proposedneeds_inputpending_approvaldonewont_do |
agent | string | Optional | Your agent id when running as part of a fleet — hides tasks another agent holds an active claim lease on, so two agents don't grab the same task |
Annotations
The hints the server sends with this tool. A host reads them to decide whether to ask you before a call.
- Read-only
readOnlyHint: true - Changes nothing.
- Destructive
destructiveHint: false - Only adds or reads; nothing already there is removed or overwritten.
- Idempotent
idempotentHint: — - Doesn't apply: the tool changes nothing.
- Open world
openWorldHint: false - Touches only your own Done account.
Which tokens can call it
A connection made with OAuth has full access. A token from Settings → Agent access → Extra MCP tokens can be narrowed, and a narrowed token never sees beyond its project or workspace.
- Full accessAllowed
- Read-only tokenAllowed
- Single-project tokenAllowed
- Single-workspace tokenAllowed
Try it
With Done connected, ask your assistant something like:
“What's in your Done queue right now?”
Connect Done to your assistant
One MCP endpoint per account, over OAuth or a token. Step-by-step setup for each client: