@aws-sdk/client-bedrock-agentcore 3.1037.0 → 3.1039.0

package/dist-types/models/models_0.d.ts

@@ -1,5 +1,5 @@
 import type { DocumentType as __DocumentType, StreamingBlobTypes } from "@smithy/types";
-import type { AutomationStreamStatus, BrowserActionStatus, BrowserEnterprisePolicyType, BrowserSessionStatus, CodeInterpreterSessionStatus, CommandExecutionStatus, ContentBlockType, DescriptorType, EventFilterCondition, ExtractionJobStatus, HarnessConversationRole, HarnessStopReason, HarnessToolType, HarnessToolUseStatus, HarnessToolUseType, LanguageRuntime, MemoryRecordStatus, MouseButton, Oauth2FlowType, OAuthGrantType, OperatorType, ProgrammingLanguage, RegistryRecordStatus, ResourceContentType, Role, ScreenshotFormat, SessionStatus, TaskStatus, ToolName } from "./enums";
+import type { ABTestExecutionStatus, ABTestStatus, AutomationStreamStatus, BatchEvaluationStatus, BrowserActionStatus, BrowserEnterprisePolicyType, BrowserSessionStatus, CloudWatchLogsFilterOperator, CodeInterpreterSessionStatus, CommandExecutionStatus, ContentBlockType, EventFilterCondition, ExtractionJobStatus, HarnessConversationRole, HarnessStopReason, HarnessToolType, HarnessToolUseStatus, HarnessToolUseType, LanguageRuntime, MemoryRecordStatus, MouseButton, Oauth2FlowType, OAuthGrantType, OperatorType, ProgrammingLanguage, RecommendationStatus, RecommendationType, ResourceContentType, Role, ScreenshotFormat, SessionStatus, TaskStatus, ToolName } from "./enums";
 import type { AccessDeniedException, ConflictException, InternalServerException, ResourceNotFoundException, RuntimeClientError, ServiceQuotaExceededException, ThrottlingException, ValidationException } from "./errors";
 /**
  * <p> The agent card definition for A2A descriptors, including the schema version and inline content that describes the agent's capabilities.</p>
@@ -28,6 +28,238 @@ export interface A2aDescriptor {
      */
     agentCard: AgentCardDefinition | undefined;
 }
+/**
+ * <p>An online evaluation configuration associated with a specific A/B test variant.</p>
+ * @public
+ */
+export interface PerVariantOnlineEvaluationConfig {
+    /**
+     * <p>The name of the variant this evaluation configuration applies to.</p>
+     * @public
+     */
+    name: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the online evaluation configuration for this variant.</p>
+     * @public
+     */
+    onlineEvaluationConfigArn: string | undefined;
+}
+/**
+ * <p>The evaluation configuration for an A/B test, specifying which online evaluation configurations to use for measuring variant performance.</p>
+ * @public
+ */
+export type ABTestEvaluationConfig = ABTestEvaluationConfig.OnlineEvaluationConfigArnMember | ABTestEvaluationConfig.PerVariantOnlineEvaluationConfigMember | ABTestEvaluationConfig.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace ABTestEvaluationConfig {
+    /**
+     * <p>The Amazon Resource Name (ARN) of a single online evaluation configuration to use for both variants.</p>
+     * @public
+     */
+    interface OnlineEvaluationConfigArnMember {
+        onlineEvaluationConfigArn: string;
+        perVariantOnlineEvaluationConfig?: never;
+        $unknown?: never;
+    }
+    /**
+     * <p>Per-variant online evaluation configurations, allowing different evaluation settings for each variant.</p>
+     * @public
+     */
+    interface PerVariantOnlineEvaluationConfigMember {
+        onlineEvaluationConfigArn?: never;
+        perVariantOnlineEvaluationConfig: PerVariantOnlineEvaluationConfig[];
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        onlineEvaluationConfigArn?: never;
+        perVariantOnlineEvaluationConfig?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        onlineEvaluationConfigArn: (value: string) => T;
+        perVariantOnlineEvaluationConfig: (value: PerVariantOnlineEvaluationConfig[]) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * <p>Statistics for the control variant in an A/B test.</p>
+ * @public
+ */
+export interface ControlStats {
+    /**
+     * <p>The name of the control variant.</p>
+     * @public
+     */
+    variantName: string | undefined;
+    /**
+     * <p>The number of sessions evaluated for the control variant.</p>
+     * @public
+     */
+    sampleSize: number | undefined;
+    /**
+     * <p>The mean evaluation score for the control variant.</p>
+     * @public
+     */
+    mean: number | undefined;
+}
+/**
+ * <p>A confidence interval for a statistical measurement.</p>
+ * @public
+ */
+export interface ConfidenceInterval {
+    /**
+     * <p>The lower bound of the confidence interval.</p>
+     * @public
+     */
+    lower?: number | undefined;
+    /**
+     * <p>The upper bound of the confidence interval.</p>
+     * @public
+     */
+    upper?: number | undefined;
+}
+/**
+ * <p>Statistical results for a treatment variant compared against the control.</p>
+ * @public
+ */
+export interface VariantResult {
+    /**
+     * <p>The name of the treatment variant.</p>
+     * @public
+     */
+    variantName: string | undefined;
+    /**
+     * <p>The number of sessions evaluated for this variant.</p>
+     * @public
+     */
+    sampleSize: number | undefined;
+    /**
+     * <p>The mean evaluation score for this variant.</p>
+     * @public
+     */
+    mean: number | undefined;
+    /**
+     * <p>The absolute change in mean score compared to the control variant.</p>
+     * @public
+     */
+    absoluteChange?: number | undefined;
+    /**
+     * <p>The percentage change in mean score compared to the control variant.</p>
+     * @public
+     */
+    percentChange?: number | undefined;
+    /**
+     * <p>The p-value indicating the statistical significance of the observed difference.</p>
+     * @public
+     */
+    pValue?: number | undefined;
+    /**
+     * <p>The confidence interval for the observed difference.</p>
+     * @public
+     */
+    confidenceInterval?: ConfidenceInterval | undefined;
+    /**
+     * <p>Whether the observed difference is statistically significant.</p>
+     * @public
+     */
+    isSignificant: boolean | undefined;
+}
+/**
+ * <p>Statistical metrics for a single evaluator comparing control and treatment variants.</p>
+ * @public
+ */
+export interface EvaluatorMetric {
+    /**
+     * <p>The Amazon Resource Name (ARN) of the evaluator.</p>
+     * @public
+     */
+    evaluatorArn: string | undefined;
+    /**
+     * <p>The statistics for the control variant.</p>
+     * @public
+     */
+    controlStats: ControlStats | undefined;
+    /**
+     * <p>The results for each treatment variant compared against the control.</p>
+     * @public
+     */
+    variantResults: VariantResult[] | undefined;
+}
+/**
+ * <p>The statistical results of an A/B test.</p>
+ * @public
+ */
+export interface ABTestResults {
+    /**
+     * <p>The timestamp when the analysis was performed.</p>
+     * @public
+     */
+    analysisTimestamp?: Date | undefined;
+    /**
+     * <p>The per-evaluator metrics comparing control and treatment variants.</p>
+     * @public
+     */
+    evaluatorMetrics: EvaluatorMetric[] | undefined;
+}
+/**
+ * <p>Summary information about an A/B test.</p>
+ * @public
+ */
+export interface ABTestSummary {
+    /**
+     * <p>The unique identifier of the A/B test.</p>
+     * @public
+     */
+    abTestId: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the A/B test.</p>
+     * @public
+     */
+    abTestArn: string | undefined;
+    /**
+     * <p>The name of the A/B test.</p>
+     * @public
+     */
+    name: string | undefined;
+    /**
+     * <p>The current status of the A/B test.</p>
+     * @public
+     */
+    status: ABTestStatus | undefined;
+    /**
+     * <p>The execution status of the A/B test.</p>
+     * @public
+     */
+    executionStatus: ABTestExecutionStatus | undefined;
+    /**
+     * <p>The description of the A/B test.</p>
+     * @public
+     */
+    description?: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the gateway used for traffic splitting.</p>
+     * @public
+     */
+    gatewayArn?: string | undefined;
+    /**
+     * <p>The timestamp when the A/B test was created.</p>
+     * @public
+     */
+    createdAt: Date | undefined;
+    /**
+     * <p>The timestamp when the A/B test was last updated.</p>
+     * @public
+     */
+    updatedAt: Date | undefined;
+}
 /**
  * <p>Contains summary information about an actor in an AgentCore Memory resource.</p>
  * @public
@@ -650,113 +882,160 @@ export interface AgentSkillsDescriptor {
     skillDefinition?: SkillDefinition | undefined;
 }
 /**
+ * <p>A value used in filter comparisons, supporting different data types.</p>
  * @public
  */
-export interface SaveBrowserSessionProfileRequest {
-    /**
-     * <p>The trace identifier for request tracking.</p>
-     * @public
-     */
-    traceId?: string | undefined;
+export type FilterValue = FilterValue.BooleanValueMember | FilterValue.DoubleValueMember | FilterValue.StringValueMember | FilterValue.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace FilterValue {
     /**
-     * <p>The parent trace information for distributed tracing.</p>
+     * <p>A string value for text-based filtering.</p>
      * @public
      */
-    traceParent?: string | undefined;
+    interface StringValueMember {
+        stringValue: string;
+        doubleValue?: never;
+        booleanValue?: never;
+        $unknown?: never;
+    }
     /**
-     * <p>The unique identifier for the browser profile. This identifier is used to reference the profile when starting new browser sessions. The identifier must follow the pattern of an alphanumeric name (up to 48 characters) followed by a hyphen and a 10-character alphanumeric suffix.</p>
+     * <p>A numeric value for numerical filtering and comparisons.</p>
      * @public
      */
-    profileIdentifier: string | undefined;
+    interface DoubleValueMember {
+        stringValue?: never;
+        doubleValue: number;
+        booleanValue?: never;
+        $unknown?: never;
+    }
     /**
-     * <p>The unique identifier of the browser associated with the session from which to save the profile.</p>
+     * <p>A boolean value for true/false filtering conditions.</p>
      * @public
      */
-    browserIdentifier: string | undefined;
+    interface BooleanValueMember {
+        stringValue?: never;
+        doubleValue?: never;
+        booleanValue: boolean;
+        $unknown?: never;
+    }
     /**
-     * <p>The unique identifier of the browser session from which to save the profile. The session must be active when saving the profile.</p>
      * @public
      */
-    sessionId: string | undefined;
+    interface $UnknownMember {
+        stringValue?: never;
+        doubleValue?: never;
+        booleanValue?: never;
+        $unknown: [string, any];
+    }
     /**
-     * <p>A unique, case-sensitive identifier to ensure that the API request completes no more than one time. If this token matches a previous request, Amazon Bedrock AgentCore ignores the request, but does not return an error.</p>
-     * @public
+     * @deprecated unused in schema-serde mode.
+     *
      */
-    clientToken?: string | undefined;
+    interface Visitor<T> {
+        stringValue: (value: string) => T;
+        doubleValue: (value: number) => T;
+        booleanValue: (value: boolean) => T;
+        _: (name: string, value: any) => T;
+    }
 }
 /**
+ * <p>A filter for narrowing down agent traces from CloudWatch Logs based on key-value comparisons.</p>
  * @public
  */
-export interface SaveBrowserSessionProfileResponse {
+export interface CloudWatchLogsFilter {
     /**
-     * <p>The unique identifier of the saved browser profile.</p>
+     * <p>The key or field name to filter on within the agent trace data.</p>
      * @public
      */
-    profileIdentifier: string | undefined;
+    key: string | undefined;
     /**
-     * <p>The unique identifier of the browser associated with the session from which the profile was saved.</p>
+     * <p>The comparison operator to use for filtering.</p>
      * @public
      */
-    browserIdentifier: string | undefined;
+    operator: CloudWatchLogsFilterOperator | undefined;
     /**
-     * <p>The unique identifier of the browser session from which the profile was saved.</p>
+     * <p>The value to compare against using the specified operator.</p>
      * @public
      */
-    sessionId: string | undefined;
+    value: FilterValue | undefined;
+}
+/**
+ * <p>A rule configuration for filtering agent traces from CloudWatch Logs.</p>
+ * @public
+ */
+export interface CloudWatchLogsRule {
     /**
-     * <p>The timestamp when the browser profile was last updated. This value is in ISO 8601 format.</p>
+     * <p>The list of filters to apply when reading agent traces.</p>
      * @public
      */
-    lastUpdatedAt: Date | undefined;
+    filters?: CloudWatchLogsFilter[] | undefined;
 }
 /**
+ * <p>Configuration for reading agent traces from CloudWatch Logs for recommendation analysis.</p>
  * @public
  */
-export interface GetBrowserSessionRequest {
+export interface CloudWatchLogsTraceConfig {
     /**
-     * <p>The unique identifier of the browser associated with the session.</p>
+     * <p>The list of CloudWatch log group ARNs to read agent traces from.</p>
      * @public
      */
-    browserIdentifier: string | undefined;
+    logGroupArns: string[] | undefined;
     /**
-     * <p>The unique identifier of the browser session to retrieve.</p>
+     * <p>The list of service names to filter traces within the specified log groups.</p>
      * @public
      */
-    sessionId: string | undefined;
-}
-/**
- * <p>The Amazon Web Services Secrets Manager location configuration.</p>
- * @public
- */
-export interface SecretsManagerLocation {
+    serviceNames: string[] | undefined;
     /**
-     * <p>The ARN of the Amazon Web Services Secrets Manager secret containing the certificate.</p>
+     * <p>The start time of the time range to read traces from.</p>
      * @public
      */
-    secretArn: string | undefined;
+    startTime: Date | undefined;
+    /**
+     * <p>The end time of the time range to read traces from.</p>
+     * @public
+     */
+    endTime: Date | undefined;
+    /**
+     * <p>Optional rule configuration for filtering traces.</p>
+     * @public
+     */
+    rule?: CloudWatchLogsRule | undefined;
 }
 /**
- * <p>The location from which to retrieve a certificate.</p>
+ * <p>The configuration specifying where to read agent traces from for recommendation analysis.</p>
  * @public
  */
-export type CertificateLocation = CertificateLocation.SecretsManagerMember | CertificateLocation.$UnknownMember;
+export type AgentTracesConfig = AgentTracesConfig.CloudwatchLogsMember | AgentTracesConfig.SessionSpansMember | AgentTracesConfig.$UnknownMember;
 /**
  * @public
  */
-export declare namespace CertificateLocation {
+export declare namespace AgentTracesConfig {
     /**
-     * <p>The Amazon Web Services Secrets Manager location of the certificate.</p>
+     * <p>Agent traces provided as inline session spans in OpenTelemetry format.</p>
      * @public
      */
-    interface SecretsManagerMember {
-        secretsManager: SecretsManagerLocation;
+    interface SessionSpansMember {
+        sessionSpans: __DocumentType[];
+        cloudwatchLogs?: never;
+        $unknown?: never;
+    }
+    /**
+     * <p>Agent traces read from CloudWatch Logs.</p>
+     * @public
+     */
+    interface CloudwatchLogsMember {
+        sessionSpans?: never;
+        cloudwatchLogs: CloudWatchLogsTraceConfig;
         $unknown?: never;
     }
     /**
      * @public
      */
     interface $UnknownMember {
-        secretsManager?: never;
+        sessionSpans?: never;
+        cloudwatchLogs?: never;
         $unknown: [string, any];
     }
     /**
@@ -764,28 +1043,148 @@ export declare namespace CertificateLocation {
      *
      */
     interface Visitor<T> {
-        secretsManager: (value: SecretsManagerLocation) => T;
+        sessionSpans: (value: __DocumentType[]) => T;
+        cloudwatchLogs: (value: CloudWatchLogsTraceConfig) => T;
         _: (name: string, value: any) => T;
     }
 }
 /**
- * <p>A certificate to install in the browser or code interpreter session.</p>
  * @public
  */
-export interface Certificate {
+export interface SaveBrowserSessionProfileRequest {
     /**
-     * <p>The location of the certificate.</p>
+     * <p>The trace identifier for request tracking.</p>
      * @public
      */
-    location: CertificateLocation | undefined;
-}
-/**
- * <p>The Amazon S3 location configuration of a resource.</p>
- * @public
- */
-export interface S3Location {
+    traceId?: string | undefined;
     /**
-     * <p>The name of the Amazon S3 bucket where the resource is stored.</p>
+     * <p>The parent trace information for distributed tracing.</p>
+     * @public
+     */
+    traceParent?: string | undefined;
+    /**
+     * <p>The unique identifier for the browser profile. This identifier is used to reference the profile when starting new browser sessions. The identifier must follow the pattern of an alphanumeric name (up to 48 characters) followed by a hyphen and a 10-character alphanumeric suffix.</p>
+     * @public
+     */
+    profileIdentifier: string | undefined;
+    /**
+     * <p>The unique identifier of the browser associated with the session from which to save the profile.</p>
+     * @public
+     */
+    browserIdentifier: string | undefined;
+    /**
+     * <p>The unique identifier of the browser session from which to save the profile. The session must be active when saving the profile.</p>
+     * @public
+     */
+    sessionId: string | undefined;
+    /**
+     * <p>A unique, case-sensitive identifier to ensure that the API request completes no more than one time. If this token matches a previous request, Amazon Bedrock AgentCore ignores the request, but does not return an error.</p>
+     * @public
+     */
+    clientToken?: string | undefined;
+}
+/**
+ * @public
+ */
+export interface SaveBrowserSessionProfileResponse {
+    /**
+     * <p>The unique identifier of the saved browser profile.</p>
+     * @public
+     */
+    profileIdentifier: string | undefined;
+    /**
+     * <p>The unique identifier of the browser associated with the session from which the profile was saved.</p>
+     * @public
+     */
+    browserIdentifier: string | undefined;
+    /**
+     * <p>The unique identifier of the browser session from which the profile was saved.</p>
+     * @public
+     */
+    sessionId: string | undefined;
+    /**
+     * <p>The timestamp when the browser profile was last updated. This value is in ISO 8601 format.</p>
+     * @public
+     */
+    lastUpdatedAt: Date | undefined;
+}
+/**
+ * @public
+ */
+export interface GetBrowserSessionRequest {
+    /**
+     * <p>The unique identifier of the browser associated with the session.</p>
+     * @public
+     */
+    browserIdentifier: string | undefined;
+    /**
+     * <p>The unique identifier of the browser session to retrieve.</p>
+     * @public
+     */
+    sessionId: string | undefined;
+}
+/**
+ * <p>The Amazon Web Services Secrets Manager location configuration.</p>
+ * @public
+ */
+export interface SecretsManagerLocation {
+    /**
+     * <p>The ARN of the Amazon Web Services Secrets Manager secret containing the certificate.</p>
+     * @public
+     */
+    secretArn: string | undefined;
+}
+/**
+ * <p>The location from which to retrieve a certificate.</p>
+ * @public
+ */
+export type CertificateLocation = CertificateLocation.SecretsManagerMember | CertificateLocation.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace CertificateLocation {
+    /**
+     * <p>The Amazon Web Services Secrets Manager location of the certificate.</p>
+     * @public
+     */
+    interface SecretsManagerMember {
+        secretsManager: SecretsManagerLocation;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        secretsManager?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        secretsManager: (value: SecretsManagerLocation) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * <p>A certificate to install in the browser or code interpreter session.</p>
+ * @public
+ */
+export interface Certificate {
+    /**
+     * <p>The location of the certificate.</p>
+     * @public
+     */
+    location: CertificateLocation | undefined;
+}
+/**
+ * <p>The Amazon S3 location configuration of a resource.</p>
+ * @public
+ */
+export interface S3Location {
+    /**
+     * <p>The name of the Amazon S3 bucket where the resource is stored.</p>
      * @public
      */
     bucket: string | undefined;
@@ -2393,114 +2792,1827 @@ export interface CompleteResourceTokenAuthRequest {
 export interface CompleteResourceTokenAuthResponse {
 }
 /**
- * <p> The input data structure containing agent session spans in OpenTelemetry format. Supports traces from frameworks like Strands (AgentCore Runtime) and LangGraph with OpenInference instrumentation for comprehensive evaluation. </p>
+ * <p>A filter to restrict which gateway target paths are included in the A/B test.</p>
  * @public
  */
-export type EvaluationInput = EvaluationInput.SessionSpansMember | EvaluationInput.$UnknownMember;
+export interface GatewayFilter {
+    /**
+     * <p>A list of target path patterns to include in the A/B test.</p>
+     * @public
+     */
+    targetPaths?: string[] | undefined;
+}
 /**
+ * <p>A reference to a specific version of a configuration bundle.</p>
  * @public
  */
-export declare namespace EvaluationInput {
+export interface ConfigurationBundleRef {
     /**
-     * <p> The collection of spans representing agent execution traces within a session. Each span contains detailed information about tool calls, model interactions, and other agent activities that can be evaluated for quality and performance. </p>
+     * <p>The Amazon Resource Name (ARN) of the configuration bundle.</p>
      * @public
      */
-    interface SessionSpansMember {
-        sessionSpans: __DocumentType[];
-        $unknown?: never;
-    }
+    bundleArn: string | undefined;
     /**
+     * <p>The version of the configuration bundle.</p>
      * @public
      */
-    interface $UnknownMember {
-        sessionSpans?: never;
-        $unknown: [string, any];
-    }
+    bundleVersion: string | undefined;
+}
+/**
+ * <p>A reference to a gateway target.</p>
+ * @public
+ */
+export interface TargetRef {
     /**
-     * @deprecated unused in schema-serde mode.
-     *
+     * <p>The name of the gateway target.</p>
+     * @public
      */
-    interface Visitor<T> {
-        sessionSpans: (value: __DocumentType[]) => T;
-        _: (name: string, value: any) => T;
-    }
+    name: string | undefined;
 }
 /**
- * <p> A content block for ground truth data in evaluation reference inputs. Supports text content for expected responses and assertions. </p>
+ * <p>The configuration for an A/B test variant.</p>
  * @public
  */
-export type EvaluationContent = EvaluationContent.TextMember | EvaluationContent.$UnknownMember;
+export interface VariantConfiguration {
+    /**
+     * <p>A reference to a configuration bundle version to use for this variant.</p>
+     * @public
+     */
+    configurationBundle?: ConfigurationBundleRef | undefined;
+    /**
+     * <p>A reference to a gateway target to route traffic to for this variant.</p>
+     * @public
+     */
+    target?: TargetRef | undefined;
+}
 /**
+ * <p>A variant in an A/B test, representing either the control (C) or treatment (T1) configuration.</p>
  * @public
  */
-export declare namespace EvaluationContent {
+export interface Variant {
     /**
-     * <p> The text content of the ground truth data. Used for expected response text and assertion statements. </p>
+     * <p>The name of the variant. Must be <code>C</code> for control or <code>T1</code> for treatment.</p>
      * @public
      */
-    interface TextMember {
-        text: string;
-        $unknown?: never;
-    }
+    name: string | undefined;
     /**
+     * <p>The percentage of traffic to route to this variant. Weights across all variants must sum to 100.</p>
      * @public
      */
-    interface $UnknownMember {
-        text?: never;
-        $unknown: [string, any];
-    }
+    weight: number | undefined;
     /**
-     * @deprecated unused in schema-serde mode.
-     *
+     * <p>The configuration for this variant, including the configuration bundle or target reference.</p>
+     * @public
      */
-    interface Visitor<T> {
-        text: (value: string) => T;
+    variantConfiguration: VariantConfiguration | undefined;
+}
+/**
+ * @public
+ */
+export interface CreateABTestRequest {
+    /**
+     * <p>The name of the A/B test. Must be unique within your account.</p>
+     * @public
+     */
+    name: string | undefined;
+    /**
+     * <p>The description of the A/B test.</p>
+     * @public
+     */
+    description?: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the gateway to use for traffic splitting.</p>
+     * @public
+     */
+    gatewayArn: string | undefined;
+    /**
+     * <p>The list of variants for the A/B test. Must contain exactly two variants: a control (C) and a treatment (T1), each with a configuration bundle or target reference and a traffic weight.</p>
+     * @public
+     */
+    variants: Variant[] | undefined;
+    /**
+     * <p>Optional filter to restrict which gateway target paths are included in the A/B test.</p>
+     * @public
+     */
+    gatewayFilter?: GatewayFilter | undefined;
+    /**
+     * <p>The evaluation configuration specifying which online evaluation configurations to use for measuring variant performance.</p>
+     * @public
+     */
+    evaluationConfig: ABTestEvaluationConfig | undefined;
+    /**
+     * <p>The IAM role ARN that grants permissions for the A/B test to access gateway and evaluation resources.</p>
+     * @public
+     */
+    roleArn: string | undefined;
+    /**
+     * <p>Whether to enable the A/B test immediately upon creation. If true, traffic splitting begins automatically.</p>
+     * @public
+     */
+    enableOnCreate?: boolean | undefined;
+    /**
+     * <p>A unique, case-sensitive identifier to ensure that the API request completes no more than one time. If this token matches a previous request, the service ignores the request, but does not return an error.</p>
+     * @public
+     */
+    clientToken?: string | undefined;
+}
+/**
+ * @public
+ */
+export interface CreateABTestResponse {
+    /**
+     * <p>The unique identifier of the created A/B test.</p>
+     * @public
+     */
+    abTestId: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the created A/B test.</p>
+     * @public
+     */
+    abTestArn: string | undefined;
+    /**
+     * <p>The name of the A/B test.</p>
+     * @public
+     */
+    name?: string | undefined;
+    /**
+     * <p>The status of the A/B test.</p>
+     * @public
+     */
+    status: ABTestStatus | undefined;
+    /**
+     * <p>The execution status indicating whether the A/B test is currently running.</p>
+     * @public
+     */
+    executionStatus: ABTestExecutionStatus | undefined;
+    /**
+     * <p>The timestamp when the A/B test was created.</p>
+     * @public
+     */
+    createdAt: Date | undefined;
+}
+/**
+ * @public
+ */
+export interface DeleteABTestRequest {
+    /**
+     * <p>The unique identifier of the A/B test to delete.</p>
+     * @public
+     */
+    abTestId: string | undefined;
+}
+/**
+ * @public
+ */
+export interface DeleteABTestResponse {
+    /**
+     * <p>The unique identifier of the deleted A/B test.</p>
+     * @public
+     */
+    abTestId: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the deleted A/B test.</p>
+     * @public
+     */
+    abTestArn: string | undefined;
+    /**
+     * <p>The status of the A/B test deletion operation.</p>
+     * @public
+     */
+    status: ABTestStatus | undefined;
+}
+/**
+ * @public
+ */
+export interface DeleteBatchEvaluationRequest {
+    /**
+     * <p>The unique identifier of the batch evaluation to delete.</p>
+     * @public
+     */
+    batchEvaluationId: string | undefined;
+}
+/**
+ * @public
+ */
+export interface DeleteBatchEvaluationResponse {
+    /**
+     * <p>The unique identifier of the deleted batch evaluation.</p>
+     * @public
+     */
+    batchEvaluationId: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the deleted batch evaluation.</p>
+     * @public
+     */
+    batchEvaluationArn: string | undefined;
+    /**
+     * <p>The status of the batch evaluation deletion operation.</p>
+     * @public
+     */
+    status: BatchEvaluationStatus | undefined;
+}
+/**
+ * @public
+ */
+export interface DeleteRecommendationRequest {
+    /**
+     * <p>The unique identifier of the recommendation to delete.</p>
+     * @public
+     */
+    recommendationId: string | undefined;
+}
+/**
+ * @public
+ */
+export interface DeleteRecommendationResponse {
+    /**
+     * <p>The unique identifier of the deleted recommendation.</p>
+     * @public
+     */
+    recommendationId: string | undefined;
+    /**
+     * <p>The status of the recommendation deletion operation.</p>
+     * @public
+     */
+    status: RecommendationStatus | undefined;
+}
+/**
+ * <p> The input data structure containing agent session spans in OpenTelemetry format. Supports traces from frameworks like Strands (AgentCore Runtime) and LangGraph with OpenInference instrumentation for comprehensive evaluation. </p>
+ * @public
+ */
+export type EvaluationInput = EvaluationInput.SessionSpansMember | EvaluationInput.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace EvaluationInput {
+    /**
+     * <p> The collection of spans representing agent execution traces within a session. Each span contains detailed information about tool calls, model interactions, and other agent activities that can be evaluated for quality and performance. </p>
+     * @public
+     */
+    interface SessionSpansMember {
+        sessionSpans: __DocumentType[];
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        sessionSpans?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        sessionSpans: (value: __DocumentType[]) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * <p> A content block for ground truth data in evaluation reference inputs. Supports text content for expected responses and assertions. </p>
+ * @public
+ */
+export type EvaluationContent = EvaluationContent.TextMember | EvaluationContent.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace EvaluationContent {
+    /**
+     * <p> The text content of the ground truth data. Used for expected response text and assertion statements. </p>
+     * @public
+     */
+    interface TextMember {
+        text: string;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        text?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        text: (value: string) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * <p> The contextual information that uniquely identifies a span within the distributed tracing system. Contains session, trace, and span identifiers used to correlate evaluation results with specific agent execution points. </p>
+ * @public
+ */
+export interface SpanContext {
+    /**
+     * <p> The unique identifier of the session containing this span. Sessions represent complete conversation flows and are detected using configurable <code>SessionTimeoutMinutes</code> (default 15 minutes). </p>
+     * @public
+     */
+    sessionId: string | undefined;
+    /**
+     * <p> The unique identifier of the trace containing this span. Traces represent individual request-response interactions within a session and group related spans together. </p>
+     * @public
+     */
+    traceId?: string | undefined;
+    /**
+     * <p> The unique identifier of the specific span being referenced. Spans represent individual operations like tool calls, model invocations, or other discrete actions within the agent's execution. </p>
+     * @public
+     */
+    spanId?: string | undefined;
+}
+/**
+ * <p> The contextual information associated with an evaluation, including span context details that identify the specific traces and sessions being evaluated within the agent's execution flow. </p>
+ * @public
+ */
+export type Context = Context.SpanContextMember | Context.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace Context {
+    /**
+     * <p> The span context information that uniquely identifies the trace and span being evaluated, including session ID, trace ID, and span ID for precise targeting within the agent's execution flow. </p>
+     * @public
+     */
+    interface SpanContextMember {
+        spanContext: SpanContext;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        spanContext?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        spanContext: (value: SpanContext) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * <p> The expected tool call trajectory for trajectory-based evaluation. </p>
+ * @public
+ */
+export interface EvaluationExpectedTrajectory {
+    /**
+     * <p> The list of tool names representing the expected tool call sequence. </p>
+     * @public
+     */
+    toolNames?: string[] | undefined;
+}
+/**
+ * <p> A reference input containing ground truth data for evaluation, scoped to a specific context level (session or trace) through its span context. </p>
+ * @public
+ */
+export interface EvaluationReferenceInput {
+    /**
+     * <p> The contextual information associated with an evaluation, including span context details that identify the specific traces and sessions being evaluated within the agent's execution flow. </p>
+     * @public
+     */
+    context: Context | undefined;
+    /**
+     * <p> The expected response for trace-level evaluation. Built-in evaluators that support this field compare the agent's actual response against this value for assessment. Custom evaluators can access it through the <code>\{expected_response\}</code> placeholder in their instructions. </p>
+     * @public
+     */
+    expectedResponse?: EvaluationContent | undefined;
+    /**
+     * <p> A list of assertion statements for session-level evaluation. Each assertion describes an expected behavior or outcome the agent should demonstrate during the session. </p>
+     * @public
+     */
+    assertions?: EvaluationContent[] | undefined;
+    /**
+     * <p> The expected tool call sequence for session-level trajectory evaluation. Contains a list of tool names representing the tools the agent is expected to invoke. </p>
+     * @public
+     */
+    expectedTrajectory?: EvaluationExpectedTrajectory | undefined;
+}
+/**
+ * <p> The specification of which trace or span IDs to evaluate within the provided input data. Allows precise targeting of evaluation at different levels: tool calls, traces, or sessions. </p>
+ * @public
+ */
+export type EvaluationTarget = EvaluationTarget.SpanIdsMember | EvaluationTarget.TraceIdsMember | EvaluationTarget.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace EvaluationTarget {
+    /**
+     * <p> The list of specific span IDs to evaluate within the provided traces. Used to target evaluation at individual tool calls or specific operations within the agent's execution flow. </p>
+     * @public
+     */
+    interface SpanIdsMember {
+        spanIds: string[];
+        traceIds?: never;
+        $unknown?: never;
+    }
+    /**
+     * <p> The list of trace IDs to evaluate, representing complete request-response interactions. Used to evaluate entire conversation turns or specific agent interactions within a session. </p>
+     * @public
+     */
+    interface TraceIdsMember {
+        spanIds?: never;
+        traceIds: string[];
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        spanIds?: never;
+        traceIds?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        spanIds: (value: string[]) => T;
+        traceIds: (value: string[]) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * @public
+ */
+export interface EvaluateRequest {
+    /**
+     * <p> The unique identifier of the evaluator to use for scoring. Can be a built-in evaluator (e.g., <code>Builtin.Helpfulness</code>, <code>Builtin.Correctness</code>) or a custom evaluator Id created through the control plane API. </p>
+     * @public
+     */
+    evaluatorId: string | undefined;
+    /**
+     * <p> The input data containing agent session spans to be evaluated. Includes a list of spans in OpenTelemetry format from supported frameworks like Strands (AgentCore Runtime) or LangGraph with OpenInference instrumentation. </p>
+     * @public
+     */
+    evaluationInput: EvaluationInput | undefined;
+    /**
+     * <p> The specific trace or span IDs to evaluate within the provided input. Allows targeting evaluation at different levels: individual tool calls, single request-response interactions (traces), or entire conversation sessions. </p>
+     * @public
+     */
+    evaluationTarget?: EvaluationTarget | undefined;
+    /**
+     * <p> Ground truth data to compare against agent responses during evaluation. Allows to provide expected responses, assertions, and expected tool trajectories at different evaluation levels. Session-level reference inputs apply to the entire conversation, while trace-level reference inputs target specific request-response interactions identified by trace ID. </p>
+     * @public
+     */
+    evaluationReferenceInputs?: EvaluationReferenceInput[] | undefined;
+}
+/**
+ * <p> The token consumption statistics for language model operations during evaluation. Provides detailed breakdown of input, output, and total tokens used for cost tracking and performance monitoring. </p>
+ * @public
+ */
+export interface TokenUsage {
+    /**
+     * <p> The number of tokens consumed for input processing during the evaluation. Includes tokens from the evaluation prompt, agent traces, and any additional context provided to the evaluator model. </p>
+     * @public
+     */
+    inputTokens?: number | undefined;
+    /**
+     * <p> The number of tokens generated by the evaluator model in its response. Includes tokens for the score, explanation, and any additional output produced during the evaluation process. </p>
+     * @public
+     */
+    outputTokens?: number | undefined;
+    /**
+     * <p> The total number of tokens consumed during the evaluation, calculated as the sum of input and output tokens. Used for cost calculation and rate limiting within the service limits. </p>
+     * @public
+     */
+    totalTokens?: number | undefined;
+}
+/**
+ * <p> The comprehensive result of an evaluation containing the score, explanation, evaluator metadata, and execution details. Provides both quantitative ratings and qualitative insights about agent performance. </p>
+ * @public
+ */
+export interface EvaluationResultContent {
+    /**
+     * <p> The Amazon Resource Name (ARN) of the evaluator used to generate this result. For custom evaluators, this is the full ARN; for built-in evaluators, this follows the pattern <code>Builtin.\{EvaluatorName\}</code>. </p>
+     * @public
+     */
+    evaluatorArn: string | undefined;
+    /**
+     * <p> The unique identifier of the evaluator that produced this result. This matches the <code>evaluatorId</code> provided in the evaluation request and can be used to identify which evaluator generated specific results. </p>
+     * @public
+     */
+    evaluatorId: string | undefined;
+    /**
+     * <p> The human-readable name of the evaluator used for this evaluation. For built-in evaluators, this is the descriptive name (e.g., "Helpfulness", "Correctness"); for custom evaluators, this is the user-defined name. </p>
+     * @public
+     */
+    evaluatorName: string | undefined;
+    /**
+     * <p> The detailed explanation provided by the evaluator describing the reasoning behind the assigned score. This qualitative feedback helps understand why specific ratings were given and provides actionable insights for improvement. </p>
+     * @public
+     */
+    explanation?: string | undefined;
+    /**
+     * <p> The contextual information associated with this evaluation result, including span context details that identify the specific traces and sessions that were evaluated. </p>
+     * @public
+     */
+    context: Context | undefined;
+    /**
+     * <p> The numerical score assigned by the evaluator according to its configured rating scale. For numerical scales, this is a decimal value within the defined range. This field is not allowed for categorical scales. </p>
+     * @public
+     */
+    value?: number | undefined;
+    /**
+     * <p> The categorical label assigned by the evaluator when using a categorical rating scale. This provides a human-readable description of the evaluation result (e.g., "Excellent", "Good", "Poor") corresponding to the numerical value. For numerical scales, this field is optional and provides a natural language explanation of what the value means (e.g., value 0.5 = "Somewhat Helpful"). </p>
+     * @public
+     */
+    label?: string | undefined;
+    /**
+     * <p> The token consumption statistics for this evaluation, including input tokens, output tokens, and total tokens used by the underlying language model during the evaluation process. </p>
+     * @public
+     */
+    tokenUsage?: TokenUsage | undefined;
+    /**
+     * <p> The error message describing what went wrong if the evaluation failed. Provides detailed information about evaluation failures to help diagnose and resolve issues with evaluator configuration or input data. </p>
+     * @public
+     */
+    errorMessage?: string | undefined;
+    /**
+     * <p> The error code indicating the type of failure that occurred during evaluation. Used to programmatically identify and handle different categories of evaluation errors. </p>
+     * @public
+     */
+    errorCode?: string | undefined;
+    /**
+     * <p> The list of reference input field names that were provided but not used by the evaluator. Helps identify which ground truth data was not consumed during evaluation. </p>
+     * @public
+     */
+    ignoredReferenceInputFields?: string[] | undefined;
+}
+/**
+ * @public
+ */
+export interface EvaluateResponse {
+    /**
+     * <p> The detailed evaluation results containing scores, explanations, and metadata. Includes the evaluator information, numerical or categorical ratings based on the evaluator's rating scale, and token usage statistics for the evaluation process. </p>
+     * @public
+     */
+    evaluationResults: EvaluationResultContent[] | undefined;
+}
+/**
+ * @public
+ */
+export interface GetABTestRequest {
+    /**
+     * <p>The unique identifier of the A/B test to retrieve.</p>
+     * @public
+     */
+    abTestId: string | undefined;
+}
+/**
+ * @public
+ */
+export interface GetABTestResponse {
+    /**
+     * <p>The unique identifier of the A/B test.</p>
+     * @public
+     */
+    abTestId: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the A/B test.</p>
+     * @public
+     */
+    abTestArn: string | undefined;
+    /**
+     * <p>The name of the A/B test.</p>
+     * @public
+     */
+    name: string | undefined;
+    /**
+     * <p>The description of the A/B test.</p>
+     * @public
+     */
+    description?: string | undefined;
+    /**
+     * <p>The current status of the A/B test.</p>
+     * @public
+     */
+    status: ABTestStatus | undefined;
+    /**
+     * <p>The execution status indicating whether the A/B test is currently running.</p>
+     * @public
+     */
+    executionStatus: ABTestExecutionStatus | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the gateway used for traffic splitting.</p>
+     * @public
+     */
+    gatewayArn: string | undefined;
+    /**
+     * <p>The list of variants in the A/B test.</p>
+     * @public
+     */
+    variants: Variant[] | undefined;
+    /**
+     * <p>The gateway filter restricting which target paths are included.</p>
+     * @public
+     */
+    gatewayFilter?: GatewayFilter | undefined;
+    /**
+     * <p>The evaluation configuration for measuring variant performance.</p>
+     * @public
+     */
+    evaluationConfig: ABTestEvaluationConfig | undefined;
+    /**
+     * <p>The IAM role ARN used by the A/B test.</p>
+     * @public
+     */
+    roleArn?: string | undefined;
+    /**
+     * <p>The identifier of the current run of the A/B test.</p>
+     * @public
+     */
+    currentRunId?: string | undefined;
+    /**
+     * <p>The error details if the A/B test encountered failures.</p>
+     * @public
+     */
+    errorDetails?: string[] | undefined;
+    /**
+     * <p>The timestamp when the A/B test was started.</p>
+     * @public
+     */
+    startedAt?: Date | undefined;
+    /**
+     * <p>The timestamp when the A/B test was stopped.</p>
+     * @public
+     */
+    stoppedAt?: Date | undefined;
+    /**
+     * <p>The timestamp when the A/B test will automatically expire.</p>
+     * @public
+     */
+    maxDurationExpiresAt?: Date | undefined;
+    /**
+     * <p>The timestamp when the A/B test was created.</p>
+     * @public
+     */
+    createdAt: Date | undefined;
+    /**
+     * <p>The timestamp when the A/B test was last updated.</p>
+     * @public
+     */
+    updatedAt: Date | undefined;
+    /**
+     * <p>The statistical results of the A/B test, including per-evaluator metrics and significance analysis.</p>
+     * @public
+     */
+    results?: ABTestResults | undefined;
+}
+/**
+ * @public
+ */
+export interface GetBatchEvaluationRequest {
+    /**
+     * <p>The unique identifier of the batch evaluation to retrieve.</p>
+     * @public
+     */
+    batchEvaluationId: string | undefined;
+}
+/**
+ * <p>A time range filter for selecting sessions. Specifies the start and end times to narrow down which sessions are included.</p>
+ * @public
+ */
+export interface SessionFilterConfig {
+    /**
+     * <p>The start time of the time range. Only sessions with activity at or after this timestamp are included.</p>
+     * @public
+     */
+    startTime?: Date | undefined;
+    /**
+     * <p>The end time of the time range. Only sessions with activity before this timestamp are included.</p>
+     * @public
+     */
+    endTime?: Date | undefined;
+}
+/**
+ * <p>Filter configuration for narrowing down CloudWatch Logs sessions for evaluation.</p>
+ * @public
+ */
+export interface CloudWatchFilterConfig {
+    /**
+     * <p>A list of specific session IDs to evaluate. If specified, only these sessions are included in the evaluation.</p>
+     * @public
+     */
+    sessionIds?: string[] | undefined;
+    /**
+     * <p>The time range filter for selecting sessions to evaluate.</p>
+     * @public
+     */
+    timeRange?: SessionFilterConfig | undefined;
+}
+/**
+ * <p>The configuration for reading agent traces from CloudWatch Logs.</p>
+ * @public
+ */
+export interface CloudWatchLogsSource {
+    /**
+     * <p>The list of agent service names to filter traces within the specified log groups.</p>
+     * @public
+     */
+    serviceNames: string[] | undefined;
+    /**
+     * <p>The list of CloudWatch log group names to read agent traces from. Maximum of 5 log groups.</p>
+     * @public
+     */
+    logGroupNames: string[] | undefined;
+    /**
+     * <p>Optional filter configuration to narrow down which sessions to evaluate.</p>
+     * @public
+     */
+    filterConfig?: CloudWatchFilterConfig | undefined;
+}
+/**
+ * Where to pull session spans from
+ * @public
+ */
+export type DataSourceConfig = DataSourceConfig.CloudWatchLogsMember | DataSourceConfig.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace DataSourceConfig {
+    /**
+     * Pull session spans from CloudWatch
+     * @public
+     */
+    interface CloudWatchLogsMember {
+        cloudWatchLogs: CloudWatchLogsSource;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        cloudWatchLogs?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        cloudWatchLogs: (value: CloudWatchLogsSource) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * <p>Aggregated statistics for an evaluator.</p>
+ * @public
+ */
+export interface EvaluatorStatistics {
+    /**
+     * <p>The average score across all evaluated sessions for this evaluator.</p>
+     * @public
+     */
+    averageScore?: number | undefined;
+}
+/**
+ * <p>Summary statistics for a single evaluator within a batch evaluation.</p>
+ * @public
+ */
+export interface EvaluatorSummary {
+    /**
+     * <p>The unique identifier of the evaluator.</p>
+     * @public
+     */
+    evaluatorId?: string | undefined;
+    /**
+     * <p>The aggregated statistics for this evaluator.</p>
+     * @public
+     */
+    statistics?: EvaluatorStatistics | undefined;
+    /**
+     * <p>The total number of sessions evaluated by this evaluator.</p>
+     * @public
+     */
+    totalEvaluated?: number | undefined;
+    /**
+     * <p>The total number of sessions that failed evaluation by this evaluator.</p>
+     * @public
+     */
+    totalFailed?: number | undefined;
+}
+/**
+ * <p>Aggregated results from a batch evaluation, including session completion counts and evaluator score summaries.</p>
+ * @public
+ */
+export interface EvaluationJobResults {
+    /**
+     * <p>The number of sessions that have been successfully evaluated.</p>
+     * @public
+     */
+    numberOfSessionsCompleted?: number | undefined;
+    /**
+     * <p>The number of sessions currently being evaluated.</p>
+     * @public
+     */
+    numberOfSessionsInProgress?: number | undefined;
+    /**
+     * <p>The number of sessions that failed evaluation.</p>
+     * @public
+     */
+    numberOfSessionsFailed?: number | undefined;
+    /**
+     * <p>The total number of sessions included in the batch evaluation.</p>
+     * @public
+     */
+    totalNumberOfSessions?: number | undefined;
+    /**
+     * <p>The number of sessions that were ignored during evaluation.</p>
+     * @public
+     */
+    numberOfSessionsIgnored?: number | undefined;
+    /**
+     * <p>A list of per-evaluator summary statistics.</p>
+     * @public
+     */
+    evaluatorSummaries?: EvaluatorSummary[] | undefined;
+}
+/**
+ * An evaluator to run against sessions
+ * @public
+ */
+export interface Evaluator {
+    /**
+     * <p>The unique identifier of the evaluator. Can reference built-in evaluators (e.g., <code>Builtin.Helpfulness</code>) or custom evaluators.</p>
+     * @public
+     */
+    evaluatorId: string | undefined;
+}
+/**
+ * CloudWatch Logs destination for batch evaluation results
+ * @public
+ */
+export interface CloudWatchOutputConfig {
+    /**
+     * <p>The name of the CloudWatch log group where evaluation results will be written.</p>
+     * @public
+     */
+    logGroupName: string | undefined;
+    /**
+     * <p>The name of the CloudWatch log stream where evaluation results will be written.</p>
+     * @public
+     */
+    logStreamName: string | undefined;
+}
+/**
+ * Output destination configuration
+ * @public
+ */
+export type OutputConfig = OutputConfig.CloudWatchConfigMember | OutputConfig.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace OutputConfig {
+    /**
+     * <p>The CloudWatch Logs configuration for writing evaluation results.</p>
+     * @public
+     */
+    interface CloudWatchConfigMember {
+        cloudWatchConfig: CloudWatchOutputConfig;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        cloudWatchConfig?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        cloudWatchConfig: (value: CloudWatchOutputConfig) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * @public
+ */
+export interface GetBatchEvaluationResponse {
+    /**
+     * <p>The unique identifier of the batch evaluation.</p>
+     * @public
+     */
+    batchEvaluationId: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the batch evaluation.</p>
+     * @public
+     */
+    batchEvaluationArn: string | undefined;
+    /**
+     * <p>The name of the batch evaluation.</p>
+     * @public
+     */
+    batchEvaluationName: string | undefined;
+    /**
+     * <p>The current status of the batch evaluation.</p>
+     * @public
+     */
+    status: BatchEvaluationStatus | undefined;
+    /**
+     * <p>The timestamp when the batch evaluation was created.</p>
+     * @public
+     */
+    createdAt: Date | undefined;
+    /**
+     * <p>The list of evaluators applied during the batch evaluation.</p>
+     * @public
+     */
+    evaluators?: Evaluator[] | undefined;
+    /**
+     * <p>The data source configuration specifying where agent traces are pulled from.</p>
+     * @public
+     */
+    dataSourceConfig?: DataSourceConfig | undefined;
+    /**
+     * <p>The output configuration specifying where evaluation results are written.</p>
+     * @public
+     */
+    outputConfig?: OutputConfig | undefined;
+    /**
+     * <p>The aggregated evaluation results, including session completion counts and evaluator score summaries.</p>
+     * @public
+     */
+    evaluationResults?: EvaluationJobResults | undefined;
+    /**
+     * <p>The error details if the batch evaluation encountered failures.</p>
+     * @public
+     */
+    errorDetails?: string[] | undefined;
+    /**
+     * <p>The description of the batch evaluation.</p>
+     * @public
+     */
+    description?: string | undefined;
+    /**
+     * <p>The timestamp when the batch evaluation was last updated.</p>
+     * @public
+     */
+    updatedAt?: Date | undefined;
+}
+/**
+ * @public
+ */
+export interface GetRecommendationRequest {
+    /**
+     * <p>The unique identifier of the recommendation to retrieve.</p>
+     * @public
+     */
+    recommendationId: string | undefined;
+}
+/**
+ * <p>A reference to an evaluator used for recommendation assessment.</p>
+ * @public
+ */
+export interface RecommendationEvaluatorReference {
+    /**
+     * <p>The Amazon Resource Name (ARN) of the evaluator.</p>
+     * @public
+     */
+    evaluatorArn: string | undefined;
+}
+/**
+ * <p>The evaluation configuration for assessing recommendation quality.</p>
+ * @public
+ */
+export interface RecommendationEvaluationConfig {
+    /**
+     * <p>The list of evaluators to use for assessing recommendation quality.</p>
+     * @public
+     */
+    evaluators: RecommendationEvaluatorReference[] | undefined;
+}
+/**
+ * <p>A system prompt sourced from a configuration bundle version.</p>
+ * @public
+ */
+export interface SystemPromptConfigurationBundle {
+    /**
+     * <p>The Amazon Resource Name (ARN) of the configuration bundle.</p>
+     * @public
+     */
+    bundleArn: string | undefined;
+    /**
+     * <p>The version identifier of the configuration bundle.</p>
+     * @public
+     */
+    versionId: string | undefined;
+    /**
+     * <p>The JSON path within the configuration bundle that contains the system prompt.</p>
+     * @public
+     */
+    systemPromptJsonPath: string | undefined;
+}
+/**
+ * <p>The system prompt input, either as inline text or from a configuration bundle.</p>
+ * @public
+ */
+export type SystemPromptConfig = SystemPromptConfig.ConfigurationBundleMember | SystemPromptConfig.TextMember | SystemPromptConfig.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace SystemPromptConfig {
+    /**
+     * <p>The system prompt text provided inline.</p>
+     * @public
+     */
+    interface TextMember {
+        text: string;
+        configurationBundle?: never;
+        $unknown?: never;
+    }
+    /**
+     * <p>The system prompt sourced from a configuration bundle version.</p>
+     * @public
+     */
+    interface ConfigurationBundleMember {
+        text?: never;
+        configurationBundle: SystemPromptConfigurationBundle;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        text?: never;
+        configurationBundle?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        text: (value: string) => T;
+        configurationBundle: (value: SystemPromptConfigurationBundle) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * <p>Configuration for generating system prompt optimization recommendations.</p>
+ * @public
+ */
+export interface SystemPromptRecommendationConfig {
+    /**
+     * <p>The current system prompt to optimize.</p>
+     * @public
+     */
+    systemPrompt: SystemPromptConfig | undefined;
+    /**
+     * <p>The agent traces to analyze for generating recommendations.</p>
+     * @public
+     */
+    agentTraces: AgentTracesConfig | undefined;
+    /**
+     * <p>The evaluation configuration specifying which evaluator to use for assessing recommendation quality.</p>
+     * @public
+     */
+    evaluationConfig: RecommendationEvaluationConfig | undefined;
+}
+/**
+ * <p>Maps a tool name to its JSON path within a configuration bundle.</p>
+ * @public
+ */
+export interface ConfigurationBundleToolEntry {
+    /**
+     * <p>The name of the tool.</p>
+     * @public
+     */
+    toolName: string | undefined;
+    /**
+     * <p>The JSON path within the configuration bundle's components that contains the tool description.</p>
+     * @public
+     */
+    toolDescriptionJsonPath: string | undefined;
+}
+/**
+ * <p>Tool descriptions sourced from a configuration bundle version.</p>
+ * @public
+ */
+export interface ToolDescriptionConfigurationBundle {
+    /**
+     * <p>The Amazon Resource Name (ARN) of the configuration bundle.</p>
+     * @public
+     */
+    bundleArn: string | undefined;
+    /**
+     * <p>The version identifier of the configuration bundle.</p>
+     * @public
+     */
+    versionId: string | undefined;
+    /**
+     * <p>The list of tool entries mapping tool names to their JSON paths within the bundle.</p>
+     * @public
+     */
+    tools: ConfigurationBundleToolEntry[] | undefined;
+}
+/**
+ * <p>The tool description content.</p>
+ * @public
+ */
+export type ToolDescriptionConfig = ToolDescriptionConfig.TextMember | ToolDescriptionConfig.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace ToolDescriptionConfig {
+    /**
+     * <p>The tool description as inline text.</p>
+     * @public
+     */
+    interface TextMember {
+        text: string;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        text?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        text: (value: string) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * <p>A tool description input containing the tool name and its current description.</p>
+ * @public
+ */
+export interface ToolDescriptionInput {
+    /**
+     * <p>The name of the tool.</p>
+     * @public
+     */
+    toolName: string | undefined;
+    /**
+     * <p>The current description of the tool to optimize.</p>
+     * @public
+     */
+    toolDescription: ToolDescriptionConfig | undefined;
+}
+/**
+ * <p>Inline tool description input containing a list of tools.</p>
+ * @public
+ */
+export interface ToolDescriptionTextInput {
+    /**
+     * <p>The list of tool descriptions to optimize.</p>
+     * @public
+     */
+    tools: ToolDescriptionInput[] | undefined;
+}
+/**
+ * <p>The source of tool descriptions, either inline text or from a configuration bundle.</p>
+ * @public
+ */
+export type ToolDescriptionSource = ToolDescriptionSource.ConfigurationBundleMember | ToolDescriptionSource.ToolDescriptionTextMember | ToolDescriptionSource.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace ToolDescriptionSource {
+    /**
+     * <p>Tool descriptions provided as inline text.</p>
+     * @public
+     */
+    interface ToolDescriptionTextMember {
+        toolDescriptionText: ToolDescriptionTextInput;
+        configurationBundle?: never;
+        $unknown?: never;
+    }
+    /**
+     * <p>Tool descriptions sourced from a configuration bundle version.</p>
+     * @public
+     */
+    interface ConfigurationBundleMember {
+        toolDescriptionText?: never;
+        configurationBundle: ToolDescriptionConfigurationBundle;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        toolDescriptionText?: never;
+        configurationBundle?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        toolDescriptionText: (value: ToolDescriptionTextInput) => T;
+        configurationBundle: (value: ToolDescriptionConfigurationBundle) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * <p>Configuration for generating tool description optimization recommendations.</p>
+ * @public
+ */
+export interface ToolDescriptionRecommendationConfig {
+    /**
+     * <p>The current tool descriptions to optimize.</p>
+     * @public
+     */
+    toolDescription: ToolDescriptionSource | undefined;
+    /**
+     * <p>The agent traces to analyze for generating tool description recommendations.</p>
+     * @public
+     */
+    agentTraces: AgentTracesConfig | undefined;
+}
+/**
+ * <p>The configuration for a recommendation, varying by recommendation type.</p>
+ * @public
+ */
+export type RecommendationConfig = RecommendationConfig.SystemPromptRecommendationConfigMember | RecommendationConfig.ToolDescriptionRecommendationConfigMember | RecommendationConfig.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace RecommendationConfig {
+    /**
+     * <p>The configuration for a system prompt recommendation.</p>
+     * @public
+     */
+    interface SystemPromptRecommendationConfigMember {
+        systemPromptRecommendationConfig: SystemPromptRecommendationConfig;
+        toolDescriptionRecommendationConfig?: never;
+        $unknown?: never;
+    }
+    /**
+     * <p>The configuration for a tool description recommendation.</p>
+     * @public
+     */
+    interface ToolDescriptionRecommendationConfigMember {
+        systemPromptRecommendationConfig?: never;
+        toolDescriptionRecommendationConfig: ToolDescriptionRecommendationConfig;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        systemPromptRecommendationConfig?: never;
+        toolDescriptionRecommendationConfig?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        systemPromptRecommendationConfig: (value: SystemPromptRecommendationConfig) => T;
+        toolDescriptionRecommendationConfig: (value: ToolDescriptionRecommendationConfig) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * <p>A configuration bundle reference in a recommendation result.</p>
+ * @public
+ */
+export interface RecommendationResultConfigurationBundle {
+    /**
+     * <p>The Amazon Resource Name (ARN) of the configuration bundle.</p>
+     * @public
+     */
+    bundleArn: string | undefined;
+    /**
+     * <p>The version identifier of the configuration bundle containing the recommendation.</p>
+     * @public
+     */
+    versionId: string | undefined;
+}
+/**
+ * <p>The result of a system prompt recommendation, containing the optimized prompt.</p>
+ * @public
+ */
+export interface SystemPromptRecommendationResult {
+    /**
+     * <p>The optimized system prompt text generated by the recommendation.</p>
+     * @public
+     */
+    recommendedSystemPrompt?: string | undefined;
+    /**
+     * <p>The configuration bundle containing the recommended system prompt, if the input was sourced from a configuration bundle.</p>
+     * @public
+     */
+    configurationBundle?: RecommendationResultConfigurationBundle | undefined;
+    /**
+     * <p>The error code if the recommendation failed.</p>
+     * @public
+     */
+    errorCode?: string | undefined;
+    /**
+     * <p>The error message if the recommendation failed.</p>
+     * @public
+     */
+    errorMessage?: string | undefined;
+}
+/**
+ * <p>The output for a single tool description recommendation.</p>
+ * @public
+ */
+export interface ToolDescriptionOutput {
+    /**
+     * <p>The name of the tool.</p>
+     * @public
+     */
+    toolName: string | undefined;
+    /**
+     * <p>The optimized tool description text generated by the recommendation.</p>
+     * @public
+     */
+    recommendedToolDescription?: string | undefined;
+}
+/**
+ * <p>The result of a tool description recommendation, containing optimized descriptions.</p>
+ * @public
+ */
+export interface ToolDescriptionRecommendationResult {
+    /**
+     * <p>The list of tools with their recommended descriptions.</p>
+     * @public
+     */
+    tools?: ToolDescriptionOutput[] | undefined;
+    /**
+     * <p>The configuration bundle containing the recommended tool descriptions, if the input was sourced from a configuration bundle.</p>
+     * @public
+     */
+    configurationBundle?: RecommendationResultConfigurationBundle | undefined;
+    /**
+     * <p>The error code if the recommendation failed.</p>
+     * @public
+     */
+    errorCode?: string | undefined;
+    /**
+     * <p>The error message if the recommendation failed.</p>
+     * @public
+     */
+    errorMessage?: string | undefined;
+}
+/**
+ * <p>The result of a recommendation, containing the optimized output.</p>
+ * @public
+ */
+export type RecommendationResult = RecommendationResult.SystemPromptRecommendationResultMember | RecommendationResult.ToolDescriptionRecommendationResultMember | RecommendationResult.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace RecommendationResult {
+    /**
+     * <p>The result of a system prompt recommendation.</p>
+     * @public
+     */
+    interface SystemPromptRecommendationResultMember {
+        systemPromptRecommendationResult: SystemPromptRecommendationResult;
+        toolDescriptionRecommendationResult?: never;
+        $unknown?: never;
+    }
+    /**
+     * <p>The result of a tool description recommendation.</p>
+     * @public
+     */
+    interface ToolDescriptionRecommendationResultMember {
+        systemPromptRecommendationResult?: never;
+        toolDescriptionRecommendationResult: ToolDescriptionRecommendationResult;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        systemPromptRecommendationResult?: never;
+        toolDescriptionRecommendationResult?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        systemPromptRecommendationResult: (value: SystemPromptRecommendationResult) => T;
+        toolDescriptionRecommendationResult: (value: ToolDescriptionRecommendationResult) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * @public
+ */
+export interface GetRecommendationResponse {
+    /**
+     * <p>The unique identifier of the recommendation.</p>
+     * @public
+     */
+    recommendationId: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the recommendation.</p>
+     * @public
+     */
+    recommendationArn: string | undefined;
+    /**
+     * <p>The name of the recommendation.</p>
+     * @public
+     */
+    name: string | undefined;
+    /**
+     * <p>The description of the recommendation.</p>
+     * @public
+     */
+    description?: string | undefined;
+    /**
+     * <p>The type of recommendation.</p>
+     * @public
+     */
+    type: RecommendationType | undefined;
+    /**
+     * <p>The configuration for the recommendation.</p>
+     * @public
+     */
+    recommendationConfig: RecommendationConfig | undefined;
+    /**
+     * <p>The current status of the recommendation.</p>
+     * @public
+     */
+    status: RecommendationStatus | undefined;
+    /**
+     * <p>The timestamp when the recommendation was created.</p>
+     * @public
+     */
+    createdAt: Date | undefined;
+    /**
+     * <p>The timestamp when the recommendation was last updated.</p>
+     * @public
+     */
+    updatedAt: Date | undefined;
+    /**
+     * <p>The result of the recommendation, containing the optimized system prompt or tool descriptions. Only present when the recommendation status is <code>COMPLETED</code>.</p>
+     * @public
+     */
+    recommendationResult?: RecommendationResult | undefined;
+}
+/**
+ * @public
+ */
+export interface ListABTestsRequest {
+    /**
+     * <p>The maximum number of results to return in the response. If the total number of results is greater than this value, use the token returned in the response in the <code>nextToken</code> field when making another request to return the next batch of results.</p>
+     * @public
+     */
+    maxResults?: number | undefined;
+    /**
+     * <p>If the total number of results is greater than the <code>maxResults</code> value provided in the request, enter the token returned in the <code>nextToken</code> field in the response in this field to return the next batch of results.</p>
+     * @public
+     */
+    nextToken?: string | undefined;
+}
+/**
+ * @public
+ */
+export interface ListABTestsResponse {
+    /**
+     * <p>The list of A/B test summaries.</p>
+     * @public
+     */
+    abTests: ABTestSummary[] | undefined;
+    /**
+     * <p>If the total number of results is greater than the <code>maxResults</code> value provided in the request, use this token when making another request in the <code>nextToken</code> field to return the next batch of results.</p>
+     * @public
+     */
+    nextToken?: string | undefined;
+}
+/**
+ * @public
+ */
+export interface ListBatchEvaluationsRequest {
+    /**
+     * <p>The maximum number of results to return in the response. If the total number of results is greater than this value, use the token returned in the response in the <code>nextToken</code> field when making another request to return the next batch of results.</p>
+     * @public
+     */
+    maxResults?: number | undefined;
+    /**
+     * <p>If the total number of results is greater than the <code>maxResults</code> value provided in the request, enter the token returned in the <code>nextToken</code> field in the response in this field to return the next batch of results.</p>
+     * @public
+     */
+    nextToken?: string | undefined;
+}
+/**
+ * Summary representation for list responses
+ * @public
+ */
+export interface BatchEvaluationSummary {
+    /**
+     * <p>The unique identifier of the batch evaluation.</p>
+     * @public
+     */
+    batchEvaluationId: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the batch evaluation.</p>
+     * @public
+     */
+    batchEvaluationArn: string | undefined;
+    /**
+     * <p>The name of the batch evaluation.</p>
+     * @public
+     */
+    batchEvaluationName: string | undefined;
+    /**
+     * <p>The current status of the batch evaluation.</p>
+     * @public
+     */
+    status: BatchEvaluationStatus | undefined;
+    /**
+     * <p>The timestamp when the batch evaluation was created.</p>
+     * @public
+     */
+    createdAt: Date | undefined;
+    /**
+     * <p>The description of the batch evaluation.</p>
+     * @public
+     */
+    description?: string | undefined;
+    /**
+     * <p>The list of evaluators applied during the batch evaluation.</p>
+     * @public
+     */
+    evaluators?: Evaluator[] | undefined;
+    /**
+     * <p>The aggregated evaluation results.</p>
+     * @public
+     */
+    evaluationResults?: EvaluationJobResults | undefined;
+    /**
+     * <p>The error details if the batch evaluation encountered failures.</p>
+     * @public
+     */
+    errorDetails?: string[] | undefined;
+    /**
+     * <p>The timestamp when the batch evaluation was last updated.</p>
+     * @public
+     */
+    updatedAt?: Date | undefined;
+}
+/**
+ * @public
+ */
+export interface ListBatchEvaluationsResponse {
+    /**
+     * <p>The list of batch evaluation summaries.</p>
+     * @public
+     */
+    batchEvaluations: BatchEvaluationSummary[] | undefined;
+    /**
+     * <p>If the total number of results is greater than the <code>maxResults</code> value provided in the request, use this token when making another request in the <code>nextToken</code> field to return the next batch of results.</p>
+     * @public
+     */
+    nextToken?: string | undefined;
+}
+/**
+ * @public
+ */
+export interface ListRecommendationsRequest {
+    /**
+     * <p>The maximum number of results to return in the response. If the total number of results is greater than this value, use the token returned in the response in the <code>nextToken</code> field when making another request to return the next batch of results.</p>
+     * @public
+     */
+    maxResults?: number | undefined;
+    /**
+     * <p>If the total number of results is greater than the <code>maxResults</code> value provided in the request, enter the token returned in the <code>nextToken</code> field in the response in this field to return the next batch of results.</p>
+     * @public
+     */
+    nextToken?: string | undefined;
+    /**
+     * <p>Optional filter to return only recommendations with the specified status.</p>
+     * @public
+     */
+    statusFilter?: RecommendationStatus | undefined;
+}
+/**
+ * <p>Summary information about a recommendation.</p>
+ * @public
+ */
+export interface RecommendationSummary {
+    /**
+     * <p>The unique identifier of the recommendation.</p>
+     * @public
+     */
+    recommendationId: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the recommendation.</p>
+     * @public
+     */
+    recommendationArn: string | undefined;
+    /**
+     * <p>The name of the recommendation.</p>
+     * @public
+     */
+    name: string | undefined;
+    /**
+     * <p>The description of the recommendation.</p>
+     * @public
+     */
+    description?: string | undefined;
+    /**
+     * <p>The type of recommendation.</p>
+     * @public
+     */
+    type: RecommendationType | undefined;
+    /**
+     * <p>The current status of the recommendation.</p>
+     * @public
+     */
+    status: RecommendationStatus | undefined;
+    /**
+     * <p>The timestamp when the recommendation was created.</p>
+     * @public
+     */
+    createdAt: Date | undefined;
+    /**
+     * <p>The timestamp when the recommendation was last updated.</p>
+     * @public
+     */
+    updatedAt: Date | undefined;
+}
+/**
+ * @public
+ */
+export interface ListRecommendationsResponse {
+    /**
+     * <p>The list of recommendation summaries.</p>
+     * @public
+     */
+    recommendationSummaries: RecommendationSummary[] | undefined;
+    /**
+     * <p>If the total number of results is greater than the <code>maxResults</code> value provided in the request, use this token when making another request in the <code>nextToken</code> field to return the next batch of results.</p>
+     * @public
+     */
+    nextToken?: string | undefined;
+}
+/**
+ * <p>The input for a ground truth conversation turn.</p>
+ * @public
+ */
+export type GroundTruthTurnInput = GroundTruthTurnInput.PromptMember | GroundTruthTurnInput.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace GroundTruthTurnInput {
+    /**
+     * <p>The text prompt for this conversation turn.</p>
+     * @public
+     */
+    interface PromptMember {
+        prompt: string;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        prompt?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        prompt: (value: string) => T;
+        _: (name: string, value: any) => T;
+    }
+}
+/**
+ * <p>Ground truth data for a single conversation turn.</p>
+ * @public
+ */
+export interface GroundTruthTurn {
+    /**
+     * <p>The input for this conversation turn.</p>
+     * @public
+     */
+    input?: GroundTruthTurnInput | undefined;
+    /**
+     * <p>The expected response for this conversation turn.</p>
+     * @public
+     */
+    expectedResponse?: EvaluationContent | undefined;
+}
+/**
+ * <p>Inline ground truth data containing assertions, expected trajectories, and per-turn expected responses.</p>
+ * @public
+ */
+export interface InlineGroundTruth {
+    /**
+     * assertions for evaluation, reuses common model EvaluationContentList
+     * @public
+     */
+    assertions?: EvaluationContent[] | undefined;
+    /**
+     * expectedTrajectory for evaluation, reuses common model EvaluationExpectedTrajectory
+     * @public
+     */
+    expectedTrajectory?: EvaluationExpectedTrajectory | undefined;
+    /**
+     * <p>A list of per-turn ground truth data, each containing an input prompt and expected response.</p>
+     * @public
+     */
+    turns?: GroundTruthTurn[] | undefined;
+}
+/**
+ * Where to pull ground truth from
+ * @public
+ */
+export type GroundTruthSource = GroundTruthSource.InlineMember | GroundTruthSource.$UnknownMember;
+/**
+ * @public
+ */
+export declare namespace GroundTruthSource {
+    /**
+     * Provide ground truth inline
+     * @public
+     */
+    interface InlineMember {
+        inline: InlineGroundTruth;
+        $unknown?: never;
+    }
+    /**
+     * @public
+     */
+    interface $UnknownMember {
+        inline?: never;
+        $unknown: [string, any];
+    }
+    /**
+     * @deprecated unused in schema-serde mode.
+     *
+     */
+    interface Visitor<T> {
+        inline: (value: InlineGroundTruth) => T;
         _: (name: string, value: any) => T;
     }
 }
 /**
- * <p> The contextual information that uniquely identifies a span within the distributed tracing system. Contains session, trace, and span identifiers used to correlate evaluation results with specific agent execution points. </p>
+ * <p>Metadata for a specific session in a batch evaluation, including ground truth data and test scenario identifiers.</p>
  * @public
  */
-export interface SpanContext {
+export interface SessionMetadataShape {
     /**
-     * <p> The unique identifier of the session containing this span. Sessions represent complete conversation flows and are detected using configurable <code>SessionTimeoutMinutes</code> (default 15 minutes). </p>
+     * <p>The unique identifier of the session this metadata applies to.</p>
      * @public
      */
     sessionId: string | undefined;
     /**
-     * <p> The unique identifier of the trace containing this span. Traces represent individual request-response interactions within a session and group related spans together. </p>
+     * <p>An optional test scenario identifier for categorizing and tracking evaluation results.</p>
      * @public
      */
-    traceId?: string | undefined;
+    testScenarioId?: string | undefined;
     /**
-     * <p> The unique identifier of the specific span being referenced. Spans represent individual operations like tool calls, model invocations, or other discrete actions within the agent's execution. </p>
+     * <p>The ground truth data for this session, including expected responses and assertions.</p>
      * @public
      */
-    spanId?: string | undefined;
+    groundTruth?: GroundTruthSource | undefined;
+    /**
+     * <p>Additional key-value metadata associated with this session.</p>
+     * @public
+     */
+    metadata?: Record<string, string> | undefined;
 }
 /**
- * <p> The contextual information associated with an evaluation, including span context details that identify the specific traces and sessions being evaluated within the agent's execution flow. </p>
+ * <p>Metadata for the evaluation, including session-specific ground truth data.</p>
  * @public
  */
-export type Context = Context.SpanContextMember | Context.$UnknownMember;
+export type EvaluationMetadata = EvaluationMetadata.SessionMetadataMember | EvaluationMetadata.$UnknownMember;
 /**
  * @public
  */
-export declare namespace Context {
+export declare namespace EvaluationMetadata {
     /**
-     * <p> The span context information that uniquely identifies the trace and span being evaluated, including session ID, trace ID, and span ID for precise targeting within the agent's execution flow. </p>
+     * <p>A list of session metadata entries containing ground truth data and test scenario identifiers for specific sessions.</p>
      * @public
      */
-    interface SpanContextMember {
-        spanContext: SpanContext;
+    interface SessionMetadataMember {
+        sessionMetadata: SessionMetadataShape[];
         $unknown?: never;
     }
     /**
      * @public
      */
     interface $UnknownMember {
-        spanContext?: never;
+        sessionMetadata?: never;
         $unknown: [string, any];
     }
     /**
@@ -2508,208 +4620,284 @@ export declare namespace Context {
      *
      */
     interface Visitor<T> {
-        spanContext: (value: SpanContext) => T;
+        sessionMetadata: (value: SessionMetadataShape[]) => T;
         _: (name: string, value: any) => T;
     }
 }
 /**
- * <p> The expected tool call trajectory for trajectory-based evaluation. </p>
  * @public
  */
-export interface EvaluationExpectedTrajectory {
+export interface StartBatchEvaluationRequest {
     /**
-     * <p> The list of tool names representing the expected tool call sequence. </p>
+     * <p>The name of the batch evaluation. Must be unique within your account.</p>
      * @public
      */
-    toolNames?: string[] | undefined;
-}
-/**
- * <p> A reference input containing ground truth data for evaluation, scoped to a specific context level (session or trace) through its span context. </p>
- * @public
- */
-export interface EvaluationReferenceInput {
+    batchEvaluationName: string | undefined;
     /**
-     * <p> The contextual information associated with an evaluation, including span context details that identify the specific traces and sessions being evaluated within the agent's execution flow. </p>
+     * <p>The list of evaluators to apply during the batch evaluation. Can include both built-in evaluators and custom evaluators. Maximum of 10 evaluators.</p>
      * @public
      */
-    context: Context | undefined;
+    evaluators?: Evaluator[] | undefined;
     /**
-     * <p> The expected response for trace-level evaluation. Built-in evaluators that support this field compare the agent's actual response against this value for assessment. Custom evaluators can access it through the <code>\{expected_response\}</code> placeholder in their instructions. </p>
+     * <p>The data source configuration that specifies where to pull agent session traces from for evaluation.</p>
      * @public
      */
-    expectedResponse?: EvaluationContent | undefined;
+    dataSourceConfig: DataSourceConfig | undefined;
     /**
-     * <p> A list of assertion statements for session-level evaluation. Each assertion describes an expected behavior or outcome the agent should demonstrate during the session. </p>
+     * <p>A unique, case-sensitive identifier to ensure that the API request completes no more than one time. If this token matches a previous request, the service ignores the request, but does not return an error.</p>
      * @public
      */
-    assertions?: EvaluationContent[] | undefined;
+    clientToken?: string | undefined;
     /**
-     * <p> The expected tool call sequence for session-level trajectory evaluation. Contains a list of tool names representing the tools the agent is expected to invoke. </p>
+     * <p>Optional metadata for the evaluation, including session-specific ground truth data and test scenario identifiers.</p>
      * @public
      */
-    expectedTrajectory?: EvaluationExpectedTrajectory | undefined;
+    evaluationMetadata?: EvaluationMetadata | undefined;
+    /**
+     * <p>The description of the batch evaluation.</p>
+     * @public
+     */
+    description?: string | undefined;
 }
-/**
- * <p> The specification of which trace or span IDs to evaluate within the provided input data. Allows precise targeting of evaluation at different levels: tool calls, traces, or sessions. </p>
- * @public
- */
-export type EvaluationTarget = EvaluationTarget.SpanIdsMember | EvaluationTarget.TraceIdsMember | EvaluationTarget.$UnknownMember;
 /**
  * @public
  */
-export declare namespace EvaluationTarget {
+export interface StartBatchEvaluationResponse {
     /**
-     * <p> The list of specific span IDs to evaluate within the provided traces. Used to target evaluation at individual tool calls or specific operations within the agent's execution flow. </p>
+     * <p>The unique identifier of the created batch evaluation.</p>
      * @public
      */
-    interface SpanIdsMember {
-        spanIds: string[];
-        traceIds?: never;
-        $unknown?: never;
-    }
+    batchEvaluationId: string | undefined;
     /**
-     * <p> The list of trace IDs to evaluate, representing complete request-response interactions. Used to evaluate entire conversation turns or specific agent interactions within a session. </p>
+     * <p>The Amazon Resource Name (ARN) of the created batch evaluation.</p>
      * @public
      */
-    interface TraceIdsMember {
-        spanIds?: never;
-        traceIds: string[];
-        $unknown?: never;
-    }
+    batchEvaluationArn: string | undefined;
     /**
+     * <p>The name of the batch evaluation.</p>
      * @public
      */
-    interface $UnknownMember {
-        spanIds?: never;
-        traceIds?: never;
-        $unknown: [string, any];
-    }
+    batchEvaluationName: string | undefined;
     /**
-     * @deprecated unused in schema-serde mode.
-     *
+     * <p>The list of evaluators applied during the batch evaluation.</p>
+     * @public
      */
-    interface Visitor<T> {
-        spanIds: (value: string[]) => T;
-        traceIds: (value: string[]) => T;
-        _: (name: string, value: any) => T;
-    }
+    evaluators?: Evaluator[] | undefined;
+    /**
+     * <p>The status of the batch evaluation.</p>
+     * @public
+     */
+    status: BatchEvaluationStatus | undefined;
+    /**
+     * <p>The timestamp when the batch evaluation was created.</p>
+     * @public
+     */
+    createdAt: Date | undefined;
+    /**
+     * <p>The output configuration specifying where evaluation results are written.</p>
+     * @public
+     */
+    outputConfig?: OutputConfig | undefined;
+    /**
+     * <p>The description of the batch evaluation.</p>
+     * @public
+     */
+    description?: string | undefined;
 }
 /**
  * @public
  */
-export interface EvaluateRequest {
+export interface StartRecommendationRequest {
     /**
-     * <p> The unique identifier of the evaluator to use for scoring. Can be a built-in evaluator (e.g., <code>Builtin.Helpfulness</code>, <code>Builtin.Correctness</code>) or a custom evaluator Id created through the control plane API. </p>
+     * <p>The name of the recommendation. Must be unique within your account.</p>
      * @public
      */
-    evaluatorId: string | undefined;
+    name: string | undefined;
     /**
-     * <p> The input data containing agent session spans to be evaluated. Includes a list of spans in OpenTelemetry format from supported frameworks like Strands (AgentCore Runtime) or LangGraph with OpenInference instrumentation. </p>
+     * <p>The description of the recommendation.</p>
      * @public
      */
-    evaluationInput: EvaluationInput | undefined;
+    description?: string | undefined;
     /**
-     * <p> The specific trace or span IDs to evaluate within the provided input. Allows targeting evaluation at different levels: individual tool calls, single request-response interactions (traces), or entire conversation sessions. </p>
+     * <p>The type of recommendation to generate. Valid values are <code>SYSTEM_PROMPT_RECOMMENDATION</code> for system prompt optimization or <code>TOOL_DESCRIPTION_RECOMMENDATION</code> for tool description optimization.</p>
      * @public
      */
-    evaluationTarget?: EvaluationTarget | undefined;
+    type: RecommendationType | undefined;
     /**
-     * <p> Ground truth data to compare against agent responses during evaluation. Allows to provide expected responses, assertions, and expected tool trajectories at different evaluation levels. Session-level reference inputs apply to the entire conversation, while trace-level reference inputs target specific request-response interactions identified by trace ID. </p>
+     * <p>The configuration for the recommendation, including the input to optimize, agent traces to analyze, and evaluation settings.</p>
      * @public
      */
-    evaluationReferenceInputs?: EvaluationReferenceInput[] | undefined;
+    recommendationConfig: RecommendationConfig | undefined;
+    /**
+     * <p>A unique, case-sensitive identifier to ensure that the API request completes no more than one time. If this token matches a previous request, the service ignores the request, but does not return an error.</p>
+     * @public
+     */
+    clientToken?: string | undefined;
 }
 /**
- * <p> The token consumption statistics for language model operations during evaluation. Provides detailed breakdown of input, output, and total tokens used for cost tracking and performance monitoring. </p>
  * @public
  */
-export interface TokenUsage {
+export interface StartRecommendationResponse {
     /**
-     * <p> The number of tokens consumed for input processing during the evaluation. Includes tokens from the evaluation prompt, agent traces, and any additional context provided to the evaluator model. </p>
+     * <p>The unique identifier of the created recommendation.</p>
      * @public
      */
-    inputTokens?: number | undefined;
+    recommendationId: string | undefined;
     /**
-     * <p> The number of tokens generated by the evaluator model in its response. Includes tokens for the score, explanation, and any additional output produced during the evaluation process. </p>
+     * <p>The Amazon Resource Name (ARN) of the created recommendation.</p>
      * @public
      */
-    outputTokens?: number | undefined;
+    recommendationArn: string | undefined;
     /**
-     * <p> The total number of tokens consumed during the evaluation, calculated as the sum of input and output tokens. Used for cost calculation and rate limiting within the service limits. </p>
+     * <p>The name of the recommendation.</p>
      * @public
      */
-    totalTokens?: number | undefined;
+    name: string | undefined;
+    /**
+     * <p>The description of the recommendation.</p>
+     * @public
+     */
+    description?: string | undefined;
+    /**
+     * <p>The type of recommendation.</p>
+     * @public
+     */
+    type: RecommendationType | undefined;
+    /**
+     * <p>The configuration for the recommendation.</p>
+     * @public
+     */
+    recommendationConfig: RecommendationConfig | undefined;
+    /**
+     * <p>The status of the recommendation.</p>
+     * @public
+     */
+    status: RecommendationStatus | undefined;
+    /**
+     * <p>The timestamp when the recommendation was created.</p>
+     * @public
+     */
+    createdAt: Date | undefined;
+    /**
+     * <p>The timestamp when the recommendation was last updated.</p>
+     * @public
+     */
+    updatedAt: Date | undefined;
 }
 /**
- * <p> The comprehensive result of an evaluation containing the score, explanation, evaluator metadata, and execution details. Provides both quantitative ratings and qualitative insights about agent performance. </p>
  * @public
  */
-export interface EvaluationResultContent {
+export interface StopBatchEvaluationRequest {
     /**
-     * <p> The Amazon Resource Name (ARN) of the evaluator used to generate this result. For custom evaluators, this is the full ARN; for built-in evaluators, this follows the pattern <code>Builtin.\{EvaluatorName\}</code>. </p>
+     * <p>The unique identifier of the batch evaluation to stop.</p>
      * @public
      */
-    evaluatorArn: string | undefined;
+    batchEvaluationId: string | undefined;
+}
+/**
+ * @public
+ */
+export interface StopBatchEvaluationResponse {
     /**
-     * <p> The unique identifier of the evaluator that produced this result. This matches the <code>evaluatorId</code> provided in the evaluation request and can be used to identify which evaluator generated specific results. </p>
+     * <p>The unique identifier of the stopped batch evaluation.</p>
      * @public
      */
-    evaluatorId: string | undefined;
+    batchEvaluationId: string | undefined;
     /**
-     * <p> The human-readable name of the evaluator used for this evaluation. For built-in evaluators, this is the descriptive name (e.g., "Helpfulness", "Correctness"); for custom evaluators, this is the user-defined name. </p>
+     * <p>The Amazon Resource Name (ARN) of the stopped batch evaluation.</p>
      * @public
      */
-    evaluatorName: string | undefined;
+    batchEvaluationArn: string | undefined;
     /**
-     * <p> The detailed explanation provided by the evaluator describing the reasoning behind the assigned score. This qualitative feedback helps understand why specific ratings were given and provides actionable insights for improvement. </p>
+     * <p>The status of the batch evaluation after the stop request.</p>
      * @public
      */
-    explanation?: string | undefined;
+    status: BatchEvaluationStatus | undefined;
     /**
-     * <p> The contextual information associated with this evaluation result, including span context details that identify the specific traces and sessions that were evaluated. </p>
+     * <p>The description of the batch evaluation.</p>
      * @public
      */
-    context: Context | undefined;
+    description?: string | undefined;
+}
+/**
+ * @public
+ */
+export interface UpdateABTestRequest {
     /**
-     * <p> The numerical score assigned by the evaluator according to its configured rating scale. For numerical scales, this is a decimal value within the defined range. This field is not allowed for categorical scales. </p>
+     * <p>The unique identifier of the A/B test to update.</p>
      * @public
      */
-    value?: number | undefined;
+    abTestId: string | undefined;
     /**
-     * <p> The categorical label assigned by the evaluator when using a categorical rating scale. This provides a human-readable description of the evaluation result (e.g., "Excellent", "Good", "Poor") corresponding to the numerical value. For numerical scales, this field is optional and provides a natural language explanation of what the value means (e.g., value 0.5 = "Somewhat Helpful"). </p>
+     * <p>A unique, case-sensitive identifier to ensure that the API request completes no more than one time. If this token matches a previous request, the service ignores the request, but does not return an error.</p>
      * @public
      */
-    label?: string | undefined;
+    clientToken?: string | undefined;
     /**
-     * <p> The token consumption statistics for this evaluation, including input tokens, output tokens, and total tokens used by the underlying language model during the evaluation process. </p>
+     * <p>The updated name of the A/B test.</p>
      * @public
      */
-    tokenUsage?: TokenUsage | undefined;
+    name?: string | undefined;
     /**
-     * <p> The error message describing what went wrong if the evaluation failed. Provides detailed information about evaluation failures to help diagnose and resolve issues with evaluator configuration or input data. </p>
+     * <p>The updated description of the A/B test.</p>
      * @public
      */
-    errorMessage?: string | undefined;
+    description?: string | undefined;
     /**
-     * <p> The error code indicating the type of failure that occurred during evaluation. Used to programmatically identify and handle different categories of evaluation errors. </p>
+     * <p>The updated list of variants.</p>
      * @public
      */
-    errorCode?: string | undefined;
+    variants?: Variant[] | undefined;
     /**
-     * <p> The list of reference input field names that were provided but not used by the evaluator. Helps identify which ground truth data was not consumed during evaluation. </p>
+     * <p>The updated gateway filter.</p>
      * @public
      */
-    ignoredReferenceInputFields?: string[] | undefined;
+    gatewayFilter?: GatewayFilter | undefined;
+    /**
+     * <p>The updated evaluation configuration.</p>
+     * @public
+     */
+    evaluationConfig?: ABTestEvaluationConfig | undefined;
+    /**
+     * <p>The updated IAM role ARN.</p>
+     * @public
+     */
+    roleArn?: string | undefined;
+    /**
+     * <p>The updated execution status to enable or disable the A/B test.</p>
+     * @public
+     */
+    executionStatus?: ABTestExecutionStatus | undefined;
 }
 /**
  * @public
  */
-export interface EvaluateResponse {
+export interface UpdateABTestResponse {
     /**
-     * <p> The detailed evaluation results containing scores, explanations, and metadata. Includes the evaluator information, numerical or categorical ratings based on the evaluator's rating scale, and token usage statistics for the evaluation process. </p>
+     * <p>The unique identifier of the updated A/B test.</p>
      * @public
      */
-    evaluationResults: EvaluationResultContent[] | undefined;
+    abTestId: string | undefined;
+    /**
+     * <p>The Amazon Resource Name (ARN) of the updated A/B test.</p>
+     * @public
+     */
+    abTestArn: string | undefined;
+    /**
+     * <p>The status of the A/B test.</p>
+     * @public
+     */
+    status: ABTestStatus | undefined;
+    /**
+     * <p>The execution status of the A/B test.</p>
+     * @public
+     */
+    executionStatus: ABTestExecutionStatus | undefined;
+    /**
+     * <p>The timestamp when the A/B test was updated.</p>
+     * @public
+     */
+    updatedAt: Date | undefined;
 }
 /**
  * @public
@@ -3906,7 +6094,7 @@ export interface HarnessAgentCoreGatewayConfig {
      */
     gatewayArn: string | undefined;
     /**
-     * <p>How Loopy authenticates to this Gateway. Defaults to AWS_IAM (SigV4) if omitted.</p>
+     * <p>How harness authenticates to this Gateway. Defaults to AWS_IAM (SigV4) if omitted.</p>
      * @public
      */
     outboundAuth?: HarnessGatewayOutboundAuth | undefined;
@@ -3938,7 +6126,7 @@ export interface HarnessRemoteMcpConfig {
      */
     url: string | undefined;
     /**
-     * Map of key/value pairs for HTTP headers.
+     * <p>Custom headers to include when connecting to the remote MCP server.</p>
      * @public
      */
     headers?: Record<string, string> | undefined;
@@ -6104,100 +8292,3 @@ export interface McpDescriptor {
      */
     tools: ToolsDefinition | undefined;
 }
-/**
- * <p> Contains the descriptor configuration for a registry record. Only the field that matches the record's <code>descriptorType</code> is populated.</p>
- * @public
- */
-export interface Descriptors {
-    /**
-     * <p> The MCP (Model Context Protocol) descriptor configuration. Populated when the record's <code>descriptorType</code> is <code>MCP</code>.</p>
-     * @public
-     */
-    mcp?: McpDescriptor | undefined;
-    /**
-     * <p> The A2A (Agent-to-Agent) descriptor configuration. Populated when the record's <code>descriptorType</code> is <code>A2A</code>.</p>
-     * @public
-     */
-    a2a?: A2aDescriptor | undefined;
-    /**
-     * <p> The custom descriptor configuration. Populated when the record's <code>descriptorType</code> is <code>CUSTOM</code>.</p>
-     * @public
-     */
-    custom?: CustomDescriptor | undefined;
-    /**
-     * <p> The agent skills descriptor configuration. Populated when the record's <code>descriptorType</code> is <code>AGENT_SKILLS</code>.</p>
-     * @public
-     */
-    agentSkills?: AgentSkillsDescriptor | undefined;
-}
-/**
- * <p> Summary information about a registry record.</p>
- * @public
- */
-export interface RegistryRecordSummary {
-    /**
-     * <p> The Amazon Resource Name (ARN) of the registry that this record belongs to.</p>
-     * @public
-     */
-    registryArn: string | undefined;
-    /**
-     * <p> The Amazon Resource Name (ARN) of the registry record.</p>
-     * @public
-     */
-    recordArn: string | undefined;
-    /**
-     * <p> The unique identifier of the registry record.</p>
-     * @public
-     */
-    recordId: string | undefined;
-    /**
-     * <p> The name of the registry record.</p>
-     * @public
-     */
-    name: string | undefined;
-    /**
-     * <p> A description of the registry record.</p>
-     * @public
-     */
-    description?: string | undefined;
-    /**
-     * <p> The type of descriptor associated with this registry record.</p>
-     * @public
-     */
-    descriptorType: DescriptorType | undefined;
-    /**
-     * <p> The descriptor configurations for this registry record.</p>
-     * @public
-     */
-    descriptors: Descriptors | undefined;
-    /**
-     * <p> The version of the registry record.</p>
-     * @public
-     */
-    version: string | undefined;
-    /**
-     * <p> The current status of the registry record.</p>
-     * @public
-     */
-    status: RegistryRecordStatus | undefined;
-    /**
-     * <p> The date and time when the registry record was created.</p>
-     * @public
-     */
-    createdAt: Date | undefined;
-    /**
-     * <p> The date and time when the registry record was last updated.</p>
-     * @public
-     */
-    updatedAt: Date | undefined;
-}
-/**
- * @public
- */
-export interface SearchRegistryRecordsResponse {
-    /**
-     * <p> The list of registry records that match the search query, ordered by relevance.</p>
-     * @public
-     */
-    registryRecords: RegistryRecordSummary[] | undefined;
-}