Get Conversation
Read a conversation back: its record, the full thread, token usage, and its sub-tasks.
conversation_kind, title, the status and result of the latest turn), the full thread with every turn’s messages and skill calls, aggregate cost and token usage, and for every sub-task the agent created: its record, its thread and its usage.
See the Conversations overview for the full conversation model.Authorizations
API Key for authentication
Response
Successful Response
Combined view of one chat conversation.
conversation_id is the parent gateway execution id, which is also the
activity-thread id. execution, activity, and usage can each be
independently None when the upstream is unavailable — the route returns
404 only when BOTH execution AND activity are missing.
usage at this root level is aggregated across the parent execution
AND every entry in sub_executions (single upstream call to the
llm_usage:aggregate endpoint). Per-sub usage is mirrored on each
SubExecutionResponse.usage so clients don't have to fan-out.
Parent gateway execution row (status, result, sub_executions, deep_planning, …).
Ordered message list (user, agent, tool calls, reasoning, …).
Aggregate cost + token usage over parent + every sub-execution. Best-effort.
Triple-shape entry for every sub-execution spawned by this conversation.
Per-message feedback rows recorded against the parent execution.

