v1

latestOpenAPI 3.1.0raw.githubusercontent.com2026-06-304917042.1 MB

evals

Get an eval run (optionally with embedded evidence)

Read one eval run. With no include the response is the shallow run+scores shape (back-compat); each include token adds the matching evidence section. Unknown tokens are ignored. Cross-org/unknown → 404.

get/v1/eval-runs/{run_id}

Path parameters

run_idstring uuid required

Query parameters

includestring nullable

comma-separated evidence to embed: session,turns,recording

Response

Successful Response

idstring uuid required

organization_idstring uuid required

session_idstring uuid required

evaluator_idstring uuid required

agent_idstring uuid nullable

channelstring nullable

statusstring required

started_atstring date-time nullable required

finished_atstring date-time nullable required

error_textstring nullable required

prompt_snapshotstring nullable required

threshold_snapshotnumber nullable required

config_snapshotobject

severity_snapshotstring required

format_snapshotstring required

is_critical_snapshotboolean

created_atstring date-time required

turns_truncatedboolean