OptionalacceptanceComputed aggregate acceptance results for this suite.
OptionalclientPhoenix client used for this suite (resolved lazily).
Call a DELETE endpoint
Call a GET endpoint
Get the Phoenix server version, returning a cached value if available.
The version is first populated from the x-phoenix-server-version
response header on any API call. If no version has been seen yet,
this method fetches GET /arize_phoenix_version to populate the cache.
Call a HEAD endpoint
Call a OPTIONS endpoint
Call a PATCH endpoint
Call a POST endpoint
Call a PUT endpoint
Call a TRACE endpoint
Unregister middleware
Register middleware
OptionaldatasetPhoenix dataset id (after upload).
OptionalevaluatorOpenInference tracer for evaluator spans.
Map of test name → resolved server-side example id. Populated after the dataset is synced.
OptionalexperimentPhoenix experiment id (created at suite start).
OptionalglobalGlobal tracer registration owned by this suite (if any).
Links printed at the end of the suite (Phoenix dataset / experiment URLs).
OptionalmaxHighest per-test repetition count seen in this suite. Used as the
experiment's repetitions value (informational; runs are keyed by
example + repetition number regardless).
The user-facing suite (and dataset/experiment) name.
OptionalprojectPhoenix project name used for the experiment.
Map of test name → registered example metadata. Filled as tests are declared inside the describe callback. Keys are test names.
Recorded test results for the reporter / summary.
OptionalsetupSetup error captured during suite initialization.
OptionaltracerOpenInference tracer for task spans.
OptionaltracerTracer provider owned by this suite (cleaned up at suite end).
Whether sync to Phoenix is disabled for this run.
OptionaltrackingReason the suite became inert (no tracking), if applicable.
OptionaluploadCount of best-effort POSTs (runs + annotations) that failed during upload.
State for a single
describe()block. Created during test collection and shared across all tests in the suite.