judges

List judge evaluations for a run

Every judge evaluation of this agent run, newest first, with scores, reasoning, status, token usage and cost.

Required permission: judges.view.

get/v1/projects/{project_id}/runs/{run_id}/judge-runs

Request

  • Base URL: https://api.connic.co
  • URL: https://api.connic.co/v1/projects/{project_id}/runs/{run_id}/judge-runs
  • Auth: HTTP bearer

Path parameters

project_idstring uuid required

ID of the project the API key belongs to.

run_idstring uuid required

Run ID from GET /runs.

Query parameters

convert_costs_to'EUR' | 'USD' nullable

Convert token costs to EUR or USD at the daily ECB reference rate of each model call

Response

Successful Response

idstring uuid required
judge_idstring uuid required

ID of the judge that ran this evaluation.

agent_run_idstring uuid required

ID of the agent run this judge evaluated.

judge_namestring nullable

Current name of the judge.

agent_namestring nullable

Agent of the evaluated run.

status'queued' | 'running' | 'completed' | 'failed' required

State of one judge evaluation: queued, running, then completed or failed.

overall_scorenumber nullable

Sum of the criterion scores; null until the evaluation completes.

overall_max_scorenumber nullable

Sum of the criteria's maximum points; null until the evaluation completes.

reasoningstring nullable

Judge's overall explanation; null until the evaluation completes and after the evaluated run is deleted.

errorstring nullable

Failure message when status is failed.

created_atstring date-time required
completed_atstring date-time nullable

When the evaluation completed or failed; null while it is queued or running.

Changes

No recorded changes to this operation across all 2 revisions of this API.