feat: Add Rubric based tool use metric

The PR does two main things:
1) Introduces a new rubric based tool use metric
2) Given that we now have two rubric based metric, we refactor and create a new RubricBasedEvaluator interface.

PiperOrigin-RevId: 811983514

Author: ankursharmas

src/google/adk/evaluation/common.py

@@ -22,6 +22,6 @@ class EvalBaseModel(pydantic.BaseModel):
   model_config = pydantic.ConfigDict(
       alias_generator=alias_generators.to_camel,
       populate_by_name=True,
-      extra='forbid',
+      extra="forbid",
       arbitrary_types_allowed=True,
   )

src/google/adk/evaluation/eval_metrics.py

@@ -52,6 +52,8 @@ class PrebuiltMetrics(Enum):
       "rubric_based_final_response_quality_v1"
   )
 
+  RUBRIC_BASED_TOOL_USE_QUALITY_V1 = "rubric_based_tool_use_quality_v1"
+
 
 MetricName: TypeAlias = Union[str, PrebuiltMetrics]
 Threshold: TypeAlias = float

src/google/adk/evaluation/llm_as_judge.py

@@ -26,6 +26,7 @@
 from ..models.llm_response import LlmResponse
 from ..models.registry import LLMRegistry
 from ..utils.context_utils import Aclosing
+from ..utils.feature_decorator import experimental
 from .common import EvalBaseModel
 from .eval_case import Invocation
 from .eval_metrics import BaseCriterion
@@ -42,6 +43,7 @@ class AutoRaterScore(EvalBaseModel):
   rubric_scores: Optional[list[RubricScore]] = None
 
 
+@experimental
 class LlmAsJudge(Evaluator):
   """Evaluator based on a LLM.
 

src/google/adk/evaluation/metric_evaluator_registry.py

@@ -25,6 +25,7 @@
 from .final_response_match_v2 import FinalResponseMatchV2Evaluator
 from .response_evaluator import ResponseEvaluator
 from .rubric_based_final_response_quality_v1 import RubricBasedFinalResponseQualityV1Evaluator
+from .rubric_based_tool_use_quality_v1 import RubricBasedToolUseV1Evaluator
 from .safety_evaluator import SafetyEvaluatorV1
 from .trajectory_evaluator import TrajectoryEvaluator
 
@@ -116,6 +117,10 @@ def _get_default_metric_evaluator_registry() -> MetricEvaluatorRegistry:
       metric_info=RubricBasedFinalResponseQualityV1Evaluator.get_metric_info(),
       evaluator=RubricBasedFinalResponseQualityV1Evaluator,
   )
+  metric_evaluator_registry.register_evaluator(
+      metric_info=RubricBasedToolUseV1Evaluator.get_metric_info(),
+      evaluator=RubricBasedToolUseV1Evaluator,
+  )
 
   return metric_evaluator_registry