Save a row's verdict
Record this cycle’s good-or-bad verdict on one row’s output. A bad verdict must carry a note saying what to change; a good one carries none. A null verdict withdraws whatever was recorded.
Authorizations
Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Body
One row's verdict, and the note a bad verdict has to carry.
What was wrong and how to change it. Required when the verdict is bad.
4000The reviewer's call on this row. Null withdraws it: an empty note no longer clears anything, because a good row is recorded without one.
good, bad Response
Successful Response
A whole training session, as the reviewer's page polls it.
Everything the page needs in one payload: the phase (status), the
cycle, the rows with their results and critiques, and the pending proposal
when there is one.
How far the current cycle's run has got.
A pending prompt revision, alongside what it would replace.
Where a session's training inputs came from.
table, manual Where a training session stands — the state the reviewer's screen shows.
One enum rather than a separate lifecycle flag and phase, because the
phase IS the state: a session is either waiting on the machine
(RUNNING), waiting on the reviewer (REVIEWING, PROPOSED), or
finished. Callers never derive the screen from row counts.
running, reviewing, proposed, completed, abandoned