Interface SendTextReply

The turn was refused by INBOUND MODERATION before any model ran (SDK 0.63.0): text is the configured refusal line, not something a model chose to say. Absent on every ordinary turn.

It exists because the two were indistinguishable on the wire — same shape, same text — so an integrator could only discover a block by A/B experiment, which is exactly what issue #3379 had to do.

Which gate refused: classifier is the moderation verdict (see blockedCategories), term is the operator's literal denylist. The matched term is deliberately never sent — this response reaches the app client, which may be the end user's own browser.

Categories that fired, on a classifier block.

WHICH classifier window fired, and its char range in the scored text. blockedCategories is a union across up to six overlapping 8,000-char windows, so on a long input it says a category fired somewhere and nothing more; this localises it.

true when the server replayed an already-completed turnId instead of running the turn (0.72.0): no model call, no billing. Absent — never false — on a turn that ran.

Typed as the companion.message envelope, so reply.envelope.payload.attachments compiles (CR-754).

The id this turn went out under (0.72.0): the turnId you passed, or the one the SDK minted. Pass it back as turnId to retry the SAME turn after an ambiguous failure — the server replays it instead of re-running.

Turn-level token totals for cost accounting (#2955), when the server reported them: whole-turn sums plus reasoningTokens (a subset of completion) and the served model when known. Absent on turns that made no billed model call.