Class BuiltinEvaluator
java.lang.Object
software.amazon.jsii.JsiiObject
software.amazon.awscdk.services.bedrockagentcore.BuiltinEvaluator
- All Implemented Interfaces:
software.amazon.jsii.JsiiSerializable
@Generated(value="jsii-pacmak/1.130.0 (build 048a5ee)",
date="2026-05-19T19:44:36.920Z")
@Stability(Stable)
public class BuiltinEvaluator
extends software.amazon.jsii.JsiiObject
Built-in evaluators provided by Amazon Bedrock AgentCore.
These evaluators assess different aspects of agent performance at various levels (session, trace, or tool call).
Example:
Evaluator customEvaluator;
OnlineEvaluationConfig evaluation = OnlineEvaluationConfig.Builder.create(this, "MixedEvaluation")
.onlineEvaluationConfigName("mixed_evaluation")
.evaluators(List.of(EvaluatorSelector.builtin(BuiltinEvaluator.HELPFULNESS), EvaluatorSelector.builtin(BuiltinEvaluator.CORRECTNESS), EvaluatorSelector.custom(customEvaluator)))
.dataSource(DataSourceConfig.fromCloudWatchLogs(CloudWatchLogsDataSourceConfig.builder()
.logGroupNames(List.of("/aws/bedrock-agentcore/my-agent"))
.serviceNames(List.of("my-agent.default"))
.build()))
.build();
-
Nested Class Summary
Nested classes/interfaces inherited from class software.amazon.jsii.JsiiObject
software.amazon.jsii.JsiiObject.InitializationMode -
Field Summary
FieldsModifier and TypeFieldDescriptionstatic final BuiltinEvaluatorEvaluates whether the response is logically structured and coherent.static final BuiltinEvaluatorEvaluates whether the response is appropriately brief without missing key information.static final BuiltinEvaluatorEvaluates whether the information in the agent's response is factually accurate.static final BuiltinEvaluatorEvaluates whether information in the response is supported by provided context/sources.static final BuiltinEvaluatorEvaluates whether the conversation successfully meets the user's goals.static final BuiltinEvaluatorEvaluates whether the response contains harmful content.static final BuiltinEvaluatorEvaluates from user's perspective how useful and valuable the agent's response is.static final BuiltinEvaluatorMeasures how well the agent follows the provided system instructions.static final BuiltinEvaluatorDetects when agent evades questions or directly refuses to answer.static final BuiltinEvaluatorEvaluates whether the response appropriately addresses the user's query.static final BuiltinEvaluatorDetects content that makes generalizations about individuals or groups.static final BuiltinEvaluatorEvaluates how accurately the agent extracts parameters from user queries.static final BuiltinEvaluatorEvaluates whether the agent selected the appropriate tool for the task. -
Constructor Summary
ConstructorsModifierConstructorDescriptionBuiltinEvaluator(String value) protectedBuiltinEvaluator(software.amazon.jsii.JsiiObject.InitializationMode initializationMode) protectedBuiltinEvaluator(software.amazon.jsii.JsiiObjectRef objRef) -
Method Summary
Methods inherited from class software.amazon.jsii.JsiiObject
jsiiAsyncCall, jsiiAsyncCall, jsiiCall, jsiiCall, jsiiGet, jsiiGet, jsiiSet, jsiiStaticCall, jsiiStaticCall, jsiiStaticGet, jsiiStaticGet, jsiiStaticSet, jsiiStaticSetMethods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, waitMethods inherited from interface software.amazon.jsii.JsiiSerializable
$jsii$toJson
-
Field Details
-
COHERENCE
Evaluates whether the response is logically structured and coherent. -
CONCISENESS
Evaluates whether the response is appropriately brief without missing key information. -
CORRECTNESS
Evaluates whether the information in the agent's response is factually accurate. -
FAITHFULNESS
Evaluates whether information in the response is supported by provided context/sources. -
GOAL_SUCCESS_RATE
Evaluates whether the conversation successfully meets the user's goals. -
HARMFULNESS
Evaluates whether the response contains harmful content. -
HELPFULNESS
Evaluates from user's perspective how useful and valuable the agent's response is. -
INSTRUCTION_FOLLOWING
Measures how well the agent follows the provided system instructions. -
REFUSAL
Detects when agent evades questions or directly refuses to answer. -
RESPONSE_RELEVANCE
Evaluates whether the response appropriately addresses the user's query. -
STEREOTYPING
Detects content that makes generalizations about individuals or groups. -
TOOL_PARAMETER_ACCURACY
Evaluates how accurately the agent extracts parameters from user queries. -
TOOL_SELECTION_ACCURACY
Evaluates whether the agent selected the appropriate tool for the task.
-
-
Constructor Details
-
BuiltinEvaluator
protected BuiltinEvaluator(software.amazon.jsii.JsiiObjectRef objRef) -
BuiltinEvaluator
protected BuiltinEvaluator(software.amazon.jsii.JsiiObject.InitializationMode initializationMode) -
BuiltinEvaluator
- Parameters:
value-- The evaluator identifier string.
-
-
Method Details
-
getValue
The string value of the built-in evaluator.
-