Skip to main content
GET
Get an evaluation run by ID

Authorizations

Authorization
string
header
required

Send "ApiKey <your_api_key>" to authenticate with an API key, or "Bearer " with a user session token.

Path Parameters

tenant_id
string
required

Tenant ID

workspace_id
string
required

Workspace ID

evaluation_run_id
string
required

Evaluation run ID (UUID)

Response

Evaluation run details

agent_evaluation_id
string
agent_evaluation_name
string
agent_id
string

Agent context, populated only when runs are listed with an evaluation_id filter (joined from the parent agent-evaluation).

agent_name
string
cancel_requested
boolean

Set by the cancel endpoint, polled by the eval worker mid-run. The worker writes the terminal 'cancelled' status; this only records the request.

created_at
string
error_kind
string

ErrorKind is the structured cause behind ErrorMessage, from the fixed vocabulary in services/aispm/eval/core/models/model_errors.py (auth_failed, unreachable, quota_exceeded, timeout, blocked, invalid_config, plan_limit_exceeded, internal_error, unknown). The same vocabulary TestConnectionResponse.Reason uses, so one UI string table serves both. Treat an unrecognised value as unknown rather than failing.

error_message
string
finished_at
string
id
string
progress
number
started_at
string
status
string
summary
object