AgentEvaluationResults Class
Definition
Important
Some information relates to prerelease product that may be substantially modified before it’s released. Microsoft makes no warranties, express or implied, with respect to the information provided here.
Aggregate evaluation results across multiple items.
public sealed class AgentEvaluationResultstype AgentEvaluationResults = classPublic NotInheritable Class AgentEvaluationResults- Inheritance
-
AgentEvaluationResults
Constructors
| Name | Description |
|---|---|
| AgentEvaluationResults(String, IEnumerable<EvaluationResult>, IReadOnlyList<EvalItem>) |
Initializes a new instance of the AgentEvaluationResults class. |
Properties
| Name | Description |
|---|---|
| AllPassed |
Gets whether all items passed. |
| DetailedItems |
Gets detailed per-item results from the Foundry output_items API, including individual evaluator scores, error info, and token usage. |
| Error |
Gets error details when the evaluation run failed. |
| EvalId |
Gets the Foundry evaluation ID (Foundry only). |
| Failed |
Gets the number of items that failed. |
| InputItems |
Gets the original eval items that produced these results, for auditing.
Each entry corresponds positionally to Items — |
| Items |
Gets the per-item MEAI evaluation results. |
| Passed |
Gets the number of items that passed. |
| PerEvaluator |
Gets per-evaluator pass/fail breakdown (Foundry only). |
| ProviderName |
Gets the evaluation provider name. |
| ReportUrl |
Gets the portal URL for viewing results (Foundry only). |
| RunId |
Gets the Foundry evaluation run ID (Foundry only). |
| Status |
Gets the evaluation run status (e.g., "completed", "failed", "canceled", "timeout"). |
| SubResults |
Gets per-agent results for workflow evaluations. |
| Total |
Gets the total number of items evaluated. |
Methods
| Name | Description |
|---|---|
| AssertAllPassed(String) |
Asserts that all items passed. Throws InvalidOperationException on failure. |
| AssertDimensionScoreAtLeast(String, Double, String, Boolean, String) |
Asserts that every item's score for the given rubric dimension is at least
|
| AssertNoFailedItems(String) |
Asserts that no item ended in a failed or errored state. Includes any sub-results from workflow evaluations. |
| AssertScoreAtLeast(Double, String, String) |
Asserts that every per-evaluator score on every item is at least |