Qué hace
Park a question for the human and MOVE ON — the one way to involve them without ending your run. Use this the moment you hit something you shouldn't decide alone: spending money, anything outward-facing (an email, a post, a reply to a customer), anything irreversible, or a genuine fork where guessing wrong is expensive. Everywhere else, decide it yourself and say what you decided in your run summary — an ask for every ambiguity turns this into the comment feed it replaced. recommendation is REQUIRED: say what you'd do absent an answer, so answering is one keystroke rather than an essay. Give 2-4 options when the answer is a choice; omit them for free text. blocking:true (the default) parks the task as needs_input and drops it out of your queue — go work something else, and read the answer next run via list_open_decisions or get_task. blocking:false leaves the task in your queue so you keep working and pick the answer up later. Only ONE open blocking question per task. This is NOT an approval and never completes anything.
Parámetros
| Nombre | Tipo | Obligatorio | Descripción |
|---|---|---|---|
id | string | Obligatorio | The task id |
question | string | Obligatorio | The question, in one sentence |
kind | string | Opcional | question = a fork; permission = may I do this outward-facing/irreversible thing; input = a value only they have Valores permitidos questionpermissioninput |
context | string | Opcional | Why you're asking and what hangs on it — what you already ruled out, and the cost of each way |
options | string[] | Opcional | 2-4 concrete choices. Omit for a free-text answer. |
recommendation | string | Obligatorio | Required, and the most useful field here — a bare label is not enough. Name the option you'd take AND why, in a sentence or two: what tipped it, and what the cost of the other way is. "Reuse existing ID — a new one re-triggers install review for every connected workspace, and the payload flag is backward compatible." beats "Reuse existing ID". Lead with the option's exact wording so the UI can highlight it. |
blocking | boolean | Opcional | True (default) parks the task until answered; false keeps working |
agent | string | Opcional | Which agent is asking |
Anotaciones
Las pistas que el servidor envía con esta herramienta. Un cliente las lee para decidir si te pregunta antes de llamarla.
- Solo lectura
readOnlyHint: false - Cambia datos de tu cuenta de Done.
- Destructiva
destructiveHint: false - Solo agrega o lee; no quita ni sobrescribe nada de lo que ya existe.
- Idempotente
idempotentHint: false - Cada llamada puede agregar algo nuevo, como un comentario, una ejecución o una tarea.
- Mundo abierto
openWorldHint: false - Solo toca tu propia cuenta de Done.
Qué tokens pueden llamarla
Una conexión hecha con OAuth tiene acceso completo. Un token de Ajustes → Acceso de agentes → Tokens de MCP adicionales se puede acotar, y un token acotado nunca ve más allá de su proyecto o espacio de trabajo.
- Acceso completoPermitido
- Token de solo lecturaRechazado
- Token de un proyectoRechazado
- Token de un espacio de trabajoPermitido
Pruébala
Con Done conectado, pídele a tu asistente algo como:
“Si necesitas que elija entre los dos proveedores, pregúntame y sigue con otro trabajo.”
Conecta Done a tu asistente
Un endpoint MCP por cuenta, con OAuth o con un token. Configuración paso a paso para cada cliente: