Mid-run questions

Let the agent pause and ask instead of guessing — the run waits, you answer, the same session continues.


Unattended agents guess; guesses compound. For steps where a wrong assumption is expensive, let the agent stop and ask you:

- type: prompt
  prompt: Clarify the requirements, then implement the task.
  allowQuestions: true

It's per-step and off by default — steps without it are steered to run unattended and never ask.

What asking looks like

When genuinely blocked, the agent asks up to four multiple-choice questions — each with a short label, two to four options with descriptions, single- or multi-select, plus an "Other" free-text escape. The run parks as awaiting_answers with the session held live; your answer resumes it in place, context intact, no restart.

Where you answer

  • Editor — the question panel docks under the live session's action bar, one question at a time.
  • Dashboard / desktop — the run page shows all questions at once; Submit answers sends them.
  • Phone — a push says "The agent needs your input"; tap it and answer from the run screen.
  • Slack — for runs born from a Slack message, a single-choice question posts native buttons right into the thread; one click answers it.

A run waiting on answers keeps its workspace alive, so it counts against the concurrency cap — unlike a run parked at a gate. Answer or cancel it rather than leaving it overnight.