OptionaldryTrue when this test ran local-only (not uploaded to Phoenix).
OptionalerrorOptionalexamplePhoenix dataset example id this run was scored against, when synced.
OptionaloutputOptionalrepetition1-based repetition index, when the test ran more than once.
OptionalrepetitionsTotal repetitions configured for this test (≥ 1).
OptionalrunPhoenix experiment run id, when the run was uploaded.
OptionaltraceOpenInference trace id for the task span, when the run was traced. Surfaced
in the reporter's failure detail so an agent can pull the exact trace from
Phoenix (e.g. px/GraphQL) to see what the task actually did.
A single test outcome surfaced to the reporter / summary printer.