chore(deps): bump the python group with 11 updates

docs/custom-evaluators.md

@@ -37,6 +37,7 @@ pip install agentevals-evaluator-sdk
 # evaluators/response_quality.py
 from agentevals_evaluator_sdk import evaluator, EvalInput, EvalResult
 
+
 @evaluator
 def response_quality(input: EvalInput) -> EvalResult:
     scores = []
@@ -53,6 +54,7 @@ def response_quality(input: EvalInput) -> EvalResult:
         per_invocation_scores=scores,
     )
 
+
 if __name__ == "__main__":
     response_quality.run()
 ```
@@ -436,7 +438,7 @@ Then register it:
 _RUNTIMES: list[Runtime] = [
     PythonRuntime(),
     NodeRuntime(),
-    GoRuntime(),       # new
+    GoRuntime(),  # new
 ]
 ```
 
@@ -484,6 +486,7 @@ To support a different evaluator registry (e.g., a custom API), implement `Evalu
 ```python
 from agentevals.evaluator.sources import EvaluatorSource, EvaluatorInfo, register_source
 
+
 class MyRegistrySource(EvaluatorSource):
     @property
     def source_name(self) -> str:
@@ -492,6 +495,7 @@ class MyRegistrySource(EvaluatorSource):
     async def list_evaluators(self) -> list[EvaluatorInfo]: ...
     async def fetch_evaluator(self, ref: str, dest: Path) -> Path: ...
 
+
 register_source(MyRegistrySource())
 ```
 

examples/README.md

@@ -69,10 +69,12 @@ For simple prompt-to-response agents, there's also a decorator shorthand:
 ```python
 app = AgentEvals(eval_set_id="my-eval")
 
+
 @app.agent
 def my_agent(prompt):
     return llm.invoke(prompt).content
 
+
 app.run(["Hello!", "Tell me a joke"])
 ```
 

examples/dice_agent/README.md

@@ -89,13 +89,8 @@ Add a new tool in `agent.py`:
 def roll_multiple(count: int, sides: int = 6) -> dict:
     """Roll multiple dice at once."""
     results = [random.randint(1, sides) for _ in range(count)]
-    return {
-        "count": count,
-        "sides": sides,
-        "results": results,
-        "total": sum(results),
-        "average": sum(results) / count
-    }
+    return {"count": count, "sides": sides, "results": results, "total": sum(results), "average": sum(results) / count}
+
 
 dice_agent = Agent(
     name="dice_agent",

examples/langchain_agent/README.md

@@ -145,7 +145,7 @@ os.environ["OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT"] = "true"
 
 ```python
 tracer_provider = TracerProvider()  # For spans
-logger_provider = LoggerProvider()   # For logs
+logger_provider = LoggerProvider()  # For logs
 ```
 
 **Both are required** - spans provide metadata, logs provide message content.
@@ -164,10 +164,12 @@ The log processor shares the WebSocket connection from the span processor.
 ```python
 loop = asyncio.new_event_loop()
 
+
 def run_loop_in_background():
     asyncio.set_event_loop(loop)
     loop.run_forever()
 
+
 thread = threading.Thread(target=run_loop_in_background, daemon=True)
 thread.start()
 ```

packages/evaluator-sdk-py/README.md

@@ -15,6 +15,7 @@ pip install agentevals-evaluator-sdk
 ```python
 from agentevals_evaluator_sdk import evaluator, EvalInput, EvalResult
 
+
 @evaluator
 def my_evaluator(input: EvalInput) -> EvalResult:
     scores = []
@@ -27,6 +28,7 @@ def my_evaluator(input: EvalInput) -> EvalResult:
         per_invocation_scores=scores,
     )
 
+
 if __name__ == "__main__":
     my_evaluator.run()
 ```