Retrieve a test result

Name of the result Example: "Test Result 1"

Name of the agent associated with this result

Current status of the result

• running - Running
• completed - Completed
• failed - Failed
• pending - Pending
• in_progress - In Progress
• evaluating - Evaluating
• in_queue - In Queue
• timeout - Timeout
• cancelled - Cancelled
• scaling_up - Scaling Up

Number of runs that fully met their expected outcomes with a score of 100

Total number of runs that had expected outcomes defined

Success rate of the test runs

Whether this test was run in text mode instead of voice mode Example: true or false

Whether this result was created by a scheduled cronjob

How this result was started: user (dashboard), api, mcp, cli, sdk, schedule or agent (the Cekura agent). Null when unknown.

Total duration of the test runs for this result Example: 22:30

Total number of test runs associated with this result Example: 10

Number of test runs that have completed successfully Example: 10

Number of test runs that were marked as successful Example: 10

Number of test runs that failed or encountered errors Example: 10

List of run IDs that got connected successfully (have transcript data). Returns empty list if no connected runs.

List of run IDs that never connected (no transcript data). Returns empty list if every run connected.

List of run IDs that failed infrastructure issues metric (score=0). Returns null if metric not found, empty list if metric exists but no failures.

List of run IDs that failed expected outcome metric (score=0). Returns null if metric not found, empty list if metric exists but no failures.

List of run IDs that completed successfully. Returns empty list if no successful runs.

List of scenario names used in the test runs for this result Example: ```

[
    {
        "id": 123,
        "name": "Scenario 1"
    },
    {
        "id": 456,
        "name": "Scenario 2"
    }
]

List of critical categories for this result Example:

[
    {
      "id": 2950,
      "name": "Pronunciation Analysis",
      "eval_type": "continuous_qualitative",
      "simulation_enabled": true,
      "observability_enabled": true
    },
    {
      "id": 3284,
      "name": "Latency",
      "eval_type": "numeric",
      "simulation_enabled": true,
      "observability_enabled": false
    },
    {
      "id": 3295,
      "name": "Detect Silence in Conversation",
      "eval_type": "binary",
      "simulation_enabled": true,
      "observability_enabled": true
    }
  ]

Generation state of the async AI summary, derived so the UI can show a reliable state instead of guessing from an empty ai_summary. completed = summary populated; skipped_no_failures = no failed runs, nothing to summarize; pending = there are failed runs but the summary isn't ready yet (still generating).

Per-rubric-rule performance breakdown. One entry per rule in the project's rubric_config: metric name, aggregated value across the result's runs, and whether that aggregate meets the rule's conditions.

Run Execution Type

Timestamp when this test result was created Example: 2021-01-01 00:00:00

Timestamp when this test result was last updated Example: 2021-01-01 00:00:00