Class: AWSCDK::BedrockAgentCore::BuiltinEvaluator
- Inherits:
-
Jsii::Object
- Object
- Jsii::Object
- AWSCDK::BedrockAgentCore::BuiltinEvaluator
- Defined in:
- bedrock_agent_core/builtin_evaluator.rb
Overview
Built-in evaluators provided by Amazon Bedrock AgentCore.
These evaluators assess different aspects of agent performance at various levels (session, trace, or tool call).
Class Method Summary collapse
-
.COHERENCE ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the response is logically structured and coherent.
-
.CONCISENESS ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the response is appropriately brief without missing key information.
-
.CORRECTNESS ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the information in the agent's response is factually accurate.
-
.FAITHFULNESS ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether information in the response is supported by provided context/sources.
-
.GOAL_SUCCESS_RATE ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the conversation successfully meets the user's goals.
-
.HARMFULNESS ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the response contains harmful content.
-
.HELPFULNESS ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates from user's perspective how useful and valuable the agent's response is.
-
.INSTRUCTION_FOLLOWING ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Measures how well the agent follows the provided system instructions.
- .jsii_overridable_methods ⇒ Object
-
.REFUSAL ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Detects when agent evades questions or directly refuses to answer.
-
.RESPONSE_RELEVANCE ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the response appropriately addresses the user's query.
-
.STEREOTYPING ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Detects content that makes generalizations about individuals or groups.
-
.TOOL_PARAMETER_ACCURACY ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates how accurately the agent extracts parameters from user queries.
-
.TOOL_SELECTION_ACCURACY ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the agent selected the appropriate tool for the task.
Instance Method Summary collapse
-
#initialize(value) ⇒ BuiltinEvaluator
constructor
A new instance of BuiltinEvaluator.
-
#value ⇒ String
The string value of the built-in evaluator.
Constructor Details
#initialize(value) ⇒ BuiltinEvaluator
Returns a new instance of BuiltinEvaluator.
12 13 14 15 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 12 def initialize(value) Jsii::Type.check_type(value, JSON.parse(Base64.strict_decode64("eyJwcmltaXRpdmUiOiJzdHJpbmcifQ==")), "value") Jsii::Object.instance_method(:initialize).bind(self).call(value) end |
Class Method Details
.COHERENCE ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the response is logically structured and coherent.
26 27 28 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 26 def self.COHERENCE() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "COHERENCE") end |
.CONCISENESS ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the response is appropriately brief without missing key information.
33 34 35 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 33 def self.CONCISENESS() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "CONCISENESS") end |
.CORRECTNESS ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the information in the agent's response is factually accurate.
40 41 42 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 40 def self.CORRECTNESS() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "CORRECTNESS") end |
.FAITHFULNESS ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether information in the response is supported by provided context/sources.
47 48 49 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 47 def self.FAITHFULNESS() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "FAITHFULNESS") end |
.GOAL_SUCCESS_RATE ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the conversation successfully meets the user's goals.
54 55 56 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 54 def self.GOAL_SUCCESS_RATE() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "GOAL_SUCCESS_RATE") end |
.HARMFULNESS ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the response contains harmful content.
61 62 63 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 61 def self.HARMFULNESS() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "HARMFULNESS") end |
.HELPFULNESS ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates from user's perspective how useful and valuable the agent's response is.
68 69 70 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 68 def self.HELPFULNESS() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "HELPFULNESS") end |
.INSTRUCTION_FOLLOWING ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Measures how well the agent follows the provided system instructions.
75 76 77 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 75 def self.INSTRUCTION_FOLLOWING() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "INSTRUCTION_FOLLOWING") end |
.jsii_overridable_methods ⇒ Object
17 18 19 20 21 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 17 def self.jsii_overridable_methods { :value => { kind: :property, name: "value", is_optional: false }, } end |
.REFUSAL ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Detects when agent evades questions or directly refuses to answer.
82 83 84 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 82 def self.REFUSAL() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "REFUSAL") end |
.RESPONSE_RELEVANCE ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the response appropriately addresses the user's query.
89 90 91 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 89 def self.RESPONSE_RELEVANCE() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "RESPONSE_RELEVANCE") end |
.STEREOTYPING ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Detects content that makes generalizations about individuals or groups.
96 97 98 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 96 def self.STEREOTYPING() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "STEREOTYPING") end |
.TOOL_PARAMETER_ACCURACY ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates how accurately the agent extracts parameters from user queries.
103 104 105 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 103 def self.TOOL_PARAMETER_ACCURACY() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "TOOL_PARAMETER_ACCURACY") end |
.TOOL_SELECTION_ACCURACY ⇒ AWSCDK::BedrockAgentCore::BuiltinEvaluator
Evaluates whether the agent selected the appropriate tool for the task.
110 111 112 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 110 def self.TOOL_SELECTION_ACCURACY() Jsii::Kernel.instance.get_static("aws-cdk-lib.aws_bedrockagentcore.BuiltinEvaluator", "TOOL_SELECTION_ACCURACY") end |
Instance Method Details
#value ⇒ String
The string value of the built-in evaluator.
117 118 119 |
# File 'bedrock_agent_core/builtin_evaluator.rb', line 117 def value() jsii_get_property("value") end |