chore(evaluator): added evaluator name and evaluator_type for report

src/strands_evals/cli/commands/validate.py

@@ -16,7 +16,7 @@ def _run(args: argparse.Namespace) -> int:
     experiment: Experiment = Experiment.from_file(args.experiment_file, custom_evaluators=custom_evaluators)
 
     case_count = len(experiment.cases)
-    evaluator_names = [evaluator.get_type_name() for evaluator in experiment.evaluators]
+    evaluator_names = [evaluator.get_name() for evaluator in experiment.evaluators]
 
     fmt = resolve_format(args)
     if fmt == "json":

src/strands_evals/evaluators/chaos/failure_communication_evaluator.py

@@ -46,8 +46,9 @@ def __init__(
         version: str = "v0",
         model: Model | str | None = None,
         system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.version = version
         default_prompt = get_template(version).SYSTEM_PROMPT
         self.system_prompt = system_prompt if system_prompt is not None else default_prompt

src/strands_evals/evaluators/chaos/partial_completion_evaluator.py

@@ -27,8 +27,9 @@ def __init__(
         version: str = "v0",
         model: Model | str | None = None,
         system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.version = version
         default_prompt = get_template(version).SYSTEM_PROMPT
         self.system_prompt = system_prompt if system_prompt is not None else default_prompt

src/strands_evals/evaluators/chaos/recovery_strategy_evaluator.py

@@ -46,8 +46,9 @@ def __init__(
         version: str = "v0",
         model: Model | str | None = None,
         system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.version = version
         default_prompt = get_template(version).SYSTEM_PROMPT
         self.system_prompt = system_prompt if system_prompt is not None else default_prompt

src/strands_evals/evaluators/coherence_evaluator.py

@@ -59,8 +59,9 @@ def __init__(
         model: Model | str | None = None,
         system_prompt: str | None = None,
         include_inputs: bool = True,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt or get_template(version).SYSTEM_PROMPT
         self.version = version
         self.model = model

src/strands_evals/evaluators/conciseness_evaluator.py

@@ -43,8 +43,9 @@ def __init__(
         model: Model | str | None = None,
         system_prompt: str | None = None,
         include_inputs: bool = True,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt or get_template(version).SYSTEM_PROMPT
         self.version = version
         self.model = model

src/strands_evals/evaluators/correctness_evaluator.py

@@ -75,8 +75,9 @@ def __init__(
         model: Model | str | None = None,
         system_prompt: str | None = None,
         reference_system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt if system_prompt is not None else get_template(version).SYSTEM_PROMPT
         self.reference_system_prompt = (
             reference_system_prompt

src/strands_evals/evaluators/deterministic/environment_state.py

@@ -13,11 +13,18 @@ def _find_state_by_name(states: list[EnvironmentState], name: str) -> Environmen
 
 
 class StateEquals(Evaluator[InputT, OutputT]):
-    """Checks if a named environment state matches an expected value."""
+    """Checks if a named environment state matches an expected value.
+
+    The `name` arg doubles as the state key in `EnvironmentState` and the
+    instance-level evaluator name surfaced through `get_name()`, so two
+    `StateEquals(name="cart")` / `StateEquals(name="balance")` instances
+    are naturally distinguishable in `EvaluationReport.cases`.
+    """
+
+    name: str  # required, not Optional — narrows the base class annotation
 
     def __init__(self, name: str, value: Any | None = None):
-        super().__init__()
-        self.name = name
+        super().__init__(name=name)
         self.value = value
 
     def evaluate(self, evaluation_case: EvaluationData[InputT, OutputT]) -> list[EvaluationOutput]:

src/strands_evals/evaluators/deterministic/output.py

@@ -7,8 +7,8 @@
 class Equals(Evaluator[InputT, OutputT]):
     """Checks if actual_output equals an expected value."""
 
-    def __init__(self, value: Any | None = None):
-        super().__init__()
+    def __init__(self, value: Any | None = None, name: str | None = None):
+        super().__init__(name=name)
         self.value = value
 
     def evaluate(self, evaluation_case: EvaluationData[InputT, OutputT]) -> list[EvaluationOutput]:
@@ -29,8 +29,8 @@ async def evaluate_async(self, evaluation_case: EvaluationData[InputT, OutputT])
 class Contains(Evaluator[InputT, OutputT]):
     """Checks if actual_output contains a substring."""
 
-    def __init__(self, value: str, case_sensitive: bool = True):
-        super().__init__()
+    def __init__(self, value: str, case_sensitive: bool = True, name: str | None = None):
+        super().__init__(name=name)
         self.value = value
         self.case_sensitive = case_sensitive
 
@@ -56,8 +56,8 @@ async def evaluate_async(self, evaluation_case: EvaluationData[InputT, OutputT])
 class StartsWith(Evaluator[InputT, OutputT]):
     """Checks if actual_output starts with a prefix."""
 
-    def __init__(self, value: str, case_sensitive: bool = True):
-        super().__init__()
+    def __init__(self, value: str, case_sensitive: bool = True, name: str | None = None):
+        super().__init__(name=name)
         self.value = value
         self.case_sensitive = case_sensitive
 

src/strands_evals/evaluators/deterministic/trajectory.py

@@ -6,8 +6,8 @@
 class ToolCalled(Evaluator[InputT, OutputT]):
     """Checks if a specific tool was called in the trajectory."""
 
-    def __init__(self, tool_name: str):
-        super().__init__()
+    def __init__(self, tool_name: str, name: str | None = None):
+        super().__init__(name=name)
         self.tool_name = tool_name
 
     def evaluate(self, evaluation_case: EvaluationData[InputT, OutputT]) -> list[EvaluationOutput]:

src/strands_evals/evaluators/evaluator.py

@@ -36,14 +36,22 @@ class Evaluator(Generic[InputT, OutputT]):
     evaluation_level: EvaluationLevel | None = None
     _trace_extractor: TraceExtractor | None = None
 
-    def __init__(self, trace_extractor: TraceExtractor | None = None):
+    def __init__(self, trace_extractor: TraceExtractor | None = None, name: str | None = None):
         """Initialize evaluator with optional custom trace extractor.
 
         Args:
             trace_extractor: Custom trace extractor. If None and evaluation_level is set,
                            a default TraceExtractor will be created.
+            name: Instance-level identifier used as the evaluator tag in
+                `EvaluationReport.cases[i]["evaluator"]` and as
+                `gen_ai.evaluation.name` on emitted spans/logs. When two
+                instances of the same class run in one experiment (e.g.,
+                `Contains(value="x")` and `Contains(value="y")`), distinct
+                names keep their results from colliding. Defaults to the
+                class name when unset.
         """
         self.aggregator = self._default_aggregator
+        self.name = name
         if trace_extractor:
             self._trace_extractor = trace_extractor
         elif self.evaluation_level:
@@ -264,6 +272,18 @@ def get_type_name(cls) -> str:
         """
         return cls.__name__
 
+    def get_name(self) -> str:
+        """Get the instance-level evaluator name, falling back to the class name.
+
+        Used for the per-row `evaluator` tag in `EvaluationReport` and the
+        `gen_ai.evaluation.name` OTel attribute. `get_type_name()` is still
+        used for class-keyed lookups such as `from_dict` registry resolution.
+
+        Returns:
+            str: The instance name if set, otherwise the class name.
+        """
+        return self.name or self.get_type_name()
+
     def to_dict(self) -> dict:
         """
         Convert the evaluator into a dictionary.

src/strands_evals/evaluators/faithfulness_evaluator.py

@@ -46,8 +46,9 @@ def __init__(
         version: str = "v0",
         model: Model | str | None = None,
         system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt if system_prompt is not None else get_template(version).SYSTEM_PROMPT
         self.version = version
         self.model = model

src/strands_evals/evaluators/goal_success_rate_evaluator.py

@@ -72,8 +72,9 @@ def __init__(
         model: Model | str | None = None,
         system_prompt: str | None = None,
         assertion_system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt if system_prompt is not None else get_template(version).SYSTEM_PROMPT
         self.assertion_system_prompt = (
             assertion_system_prompt

src/strands_evals/evaluators/harmfulness_evaluator.py

@@ -40,8 +40,9 @@ def __init__(
         version: str = "v0",
         model: Model | str | None = None,
         system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt if system_prompt is not None else get_template(version).SYSTEM_PROMPT
         self.version = version
         self.model = model

src/strands_evals/evaluators/helpfulness_evaluator.py

@@ -51,8 +51,9 @@ def __init__(
         model: Model | str | None = None,
         system_prompt: str | None = None,
         include_inputs: bool = True,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt if system_prompt is not None else get_template(version).SYSTEM_PROMPT
         self.version = version
         self.model = model

src/strands_evals/evaluators/instruction_following_evaluator.py

@@ -40,8 +40,9 @@ def __init__(
         version: str = "v0",
         model: Model | str | None = None,
         system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt if system_prompt is not None else get_template(version).SYSTEM_PROMPT
         self.version = version
         self.model = model

src/strands_evals/evaluators/interactions_evaluator.py

@@ -32,8 +32,9 @@ def __init__(
         model: Model | str | None = None,
         system_prompt: str = SYSTEM_PROMPT,
         include_inputs: bool = True,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.rubric = rubric
         self.interaction_description = interaction_description
         self.model = model

src/strands_evals/evaluators/multimodal_correctness_evaluator.py

@@ -24,6 +24,7 @@ def __init__(
         system_prompt: str | None = None,
         reference_suffix: str | None = None,
         uses_environment_state: bool = False,
+        name: str | None = None,
     ):
         super().__init__(
             rubric=rubric if rubric is not None else CORRECTNESS_RUBRIC_V0,
@@ -32,4 +33,5 @@ def __init__(
             system_prompt=system_prompt,
             reference_suffix=reference_suffix,
             uses_environment_state=uses_environment_state,
+            name=name,
         )

src/strands_evals/evaluators/multimodal_faithfulness_evaluator.py

@@ -23,6 +23,7 @@ def __init__(
         system_prompt: str | None = None,
         reference_suffix: str | None = None,
         uses_environment_state: bool = False,
+        name: str | None = None,
     ):
         super().__init__(
             rubric=rubric if rubric is not None else FAITHFULNESS_RUBRIC_V0,
@@ -31,4 +32,5 @@ def __init__(
             system_prompt=system_prompt,
             reference_suffix=reference_suffix,
             uses_environment_state=uses_environment_state,
+            name=name,
         )

src/strands_evals/evaluators/multimodal_instruction_following_evaluator.py

@@ -23,6 +23,7 @@ def __init__(
         system_prompt: str | None = None,
         reference_suffix: str | None = None,
         uses_environment_state: bool = False,
+        name: str | None = None,
     ):
         super().__init__(
             rubric=rubric if rubric is not None else INSTRUCTION_FOLLOWING_RUBRIC_V0,
@@ -31,4 +32,5 @@ def __init__(
             system_prompt=system_prompt,
             reference_suffix=reference_suffix,
             uses_environment_state=uses_environment_state,
+            name=name,
         )

src/strands_evals/evaluators/multimodal_output_evaluator.py

@@ -45,13 +45,15 @@ def __init__(
         system_prompt: str | None = None,
         reference_suffix: str | None = None,
         uses_environment_state: bool = False,
+        name: str | None = None,
     ):
         super().__init__(
             rubric=rubric,
             model=model,
             system_prompt=system_prompt if system_prompt is not None else MLLM_JUDGE_SYSTEM_PROMPT,
             include_inputs=include_inputs,
             uses_environment_state=uses_environment_state,
+            name=name,
         )
         self.reference_suffix = reference_suffix if reference_suffix is not None else self.DEFAULT_REFERENCE_SUFFIX
 

src/strands_evals/evaluators/multimodal_overall_quality_evaluator.py

@@ -34,6 +34,7 @@ def __init__(
         system_prompt: str | None = None,
         reference_suffix: str | None = None,
         uses_environment_state: bool = False,
+        name: str | None = None,
     ):
         super().__init__(
             rubric=rubric if rubric is not None else OVERALL_QUALITY_RUBRIC_V0,
@@ -42,4 +43,5 @@ def __init__(
             system_prompt=system_prompt,
             reference_suffix=reference_suffix if reference_suffix is not None else _OVERALL_QUALITY_REFERENCE_SUFFIX,
             uses_environment_state=uses_environment_state,
+            name=name,
         )

src/strands_evals/evaluators/output_evaluator.py

@@ -29,8 +29,9 @@ def __init__(
         system_prompt: str = SYSTEM_PROMPT,
         include_inputs: bool = True,
         uses_environment_state: bool = False,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.rubric = rubric
         self.model = model
         self.include_inputs = include_inputs

src/strands_evals/evaluators/refusal_evaluator.py

@@ -40,8 +40,9 @@ def __init__(
         version: str = "v0",
         model: Model | str | None = None,
         system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt if system_prompt is not None else get_template(version).SYSTEM_PROMPT
         self.version = version
         self.model = model

src/strands_evals/evaluators/response_relevance_evaluator.py

@@ -48,8 +48,9 @@ def __init__(
         model: Model | str | None = None,
         system_prompt: str | None = None,
         include_inputs: bool = True,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt or get_template(version).SYSTEM_PROMPT
         self.version = version
         self.model = model

src/strands_evals/evaluators/stereotyping_evaluator.py

@@ -40,8 +40,9 @@ def __init__(
         version: str = "v0",
         model: Model | str | None = None,
         system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt if system_prompt is not None else get_template(version).SYSTEM_PROMPT
         self.version = version
         self.model = model

src/strands_evals/evaluators/tool_parameter_accuracy_evaluator.py

@@ -40,8 +40,9 @@ def __init__(
         version: str = "v0",
         model: Model | str | None = None,
         system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt if system_prompt is not None else get_template(version).SYSTEM_PROMPT
         self.version = version
         self.model = model

src/strands_evals/evaluators/tool_selection_accuracy_evaluator.py

@@ -40,8 +40,9 @@ def __init__(
         version: str = "v0",
         model: Model | str | None = None,
         system_prompt: str | None = None,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.system_prompt = system_prompt if system_prompt is not None else get_template(version).SYSTEM_PROMPT
         self.version = version
         self.model = model

src/strands_evals/evaluators/trajectory_evaluator.py

@@ -32,8 +32,9 @@ def __init__(
         model: Model | str | None = None,
         system_prompt: str = SYSTEM_PROMPT,
         include_inputs: bool = True,
+        name: str | None = None,
     ):
-        super().__init__()
+        super().__init__(name=name)
         self.rubric = rubric
         self.trajectory_description = trajectory_description
         self.model = model