--timeout to wait for Mo to finish its
current turn, ask a question, or reach the time limit. Use
qa wait when a script needs to wait without
a time limit.
Arguments
Options
Maximum time to wait for new output. Accepts
0 (return immediately) or a
value such as 30s, 2m, or 500ms, up to 290s. Loop the command to
follow a long session.Where to read from.
start returns the whole transcript plus anything Mo says
while the command waits. latest skips the existing transcript and returns
only output produced while the command waits, so a polling loop does not
reprint earlier messages.Output
In plain-text mode, the command prints each message on stdout. One status message can appear on stderr:Mo was stopped.when the session was stopped, for example afterqa stopor a stop from the dashboard.Mo has N open questions:(orMo has 1 open question:) when Mo is waiting for answers, followed by one- <id>: <question> Options: a | b.line per card and theqa send <session-id> --question-id <id> "<answer>"command to answer each.Mo is still working.when the timeout ran out before Mo finished.
--json, the object has this shape:
status is the session’s state as the UI shows it, the same word
qa status reports: running,
waiting_on_agents, checking (Mo’s last outcome is unknown), needs_you,
blocked, ready (Mo marked the task complete), sleeping (Mo stopped without
marking it complete), cancelled (someone stopped the session), or failed.
Servers running an earlier API omit status; use qa status for it on those.
completed, when present, is the root agent’s explicit completion flag. A reply
or a finished turn alone does not mean the overall task is complete. timedOut
is true when the command returned because the timeout ran out while Mo was
still working. messages entries have a role of user or assistant,
text, and an optional stable id. Older servers omit id. pendingQuestions
lists every open question card, oldest first, and pendingInput is the oldest;
both are absent when no card is open. Reply with
qa send; a plain send answers the oldest
card, and --question-id <id> answers a specific one.
Examples
Follow a session until Mo stops working:Notes
Reading does not remove messages from the session. For a snapshot of the session state and its findings, runqa status
instead.