Workflows Eval Scores List

GET

https://api.respan.ai/clickhouse/workflows/:workflow_id/eval-scores

Output-score aggregate for one evaluator pipeline, keyed by family id.

One row — the pipeline’s output (rollup) score, not each grader. Numeric output uses avg_primary_score, boolean uses true_ratio. The FE plots it as the one line on the scores-over-time chart. With scope=automation the family id is an automation’s and rows match on the stamped caller automation_id — only that automation’s scores, across every evaluator pipeline it runs.

Output-score aggregate for one evaluator pipeline, keyed by family id. One row — the pipeline's output (rollup) score, not each grader. Numeric output uses ``avg_primary_score``, boolean uses ``true_ratio``. The FE plots it as the one line on the scores-over-time chart. With ``scope=automation`` the family id is an automation's and rows match on the stamped caller ``automation_id`` — only that automation's scores, across every evaluator pipeline it runs.

Authentication

AuthorizationBearer

JWT access token or Respan API key

Path parameters

workflow_idstringRequired

Query parameters

scopeenumOptional

Set to ‘automation’ when the URL family id is an automation’s — scores are then matched by the caller automation_id stamped on each row instead of the evaluator pipeline’s version ids.

Allowed values:

Response

workflow_version_idstring

evaluator_idstring

scorerstring

evaluator_namestring

sample_countinteger

avg_primary_scoredouble or null

true_ratiodouble or null

1	[
2	{
3	"workflow_version_id": "string",
4	"evaluator_id": "string",
5	"scorer": "string",
6	"evaluator_name": "string",
7	"sample_count": 1,
8	"avg_primary_score": 1.1,
9	"true_ratio": 1.1
10	}
11	]