Add py.typed marker for type checker support

README.md

@@ -203,6 +203,21 @@ poetry run python3 -m pytest tests -v
 
 Latest verification (2025-10-24): 144 passed, 10 skipped, 32 warnings. Warnings originate from third-party dependencies (`langchain` pydantic shim deprecations and `datetime.utcnow` usage). Track upstream fixes or pin patched releases as needed.
 
+### Running CI Checks Locally
+
+Before opening a pull request, you can run the same checks locally that are executed in CI.
+
+#### Requirements
+- Python **3.10 or higher**
+- [Poetry](https://python-poetry.org/) installed
+
+#### Setup
+Install dependencies (including dev tools):
+
+```bash
+poetry install --with dev
+```
+
 ## Contributing
 
 We welcome contributions! Please see [CONTRIBUTING.md](CONTRIBUTING.md) for:

src/agentunit/adapters/base.py

@@ -32,14 +32,38 @@ class BaseAdapter(abc.ABC):
 
     @abc.abstractmethod
     def prepare(self) -> None:
-        """Perform any lazy setup (loading graphs, flows, etc.)."""
+        """
+        Perform any lazy setup required before execution.
+
+        This may include loading graphs, flows, or other resources.
+
+        Returns:
+            None
+        """
 
     @abc.abstractmethod
     def execute(self, case: DatasetCase, trace: TraceLog) -> AdapterOutcome:
-        """Run the agent flow on a single dataset case."""
+        """
+        Run the agent flow on a single dataset case.
+
+        Args:
+            case (DatasetCase): The dataset case to be processed.
+            trace (TraceLog): Trace log used to record execution details.
+
+        Returns:
+            AdapterOutcome: The outcome produced by executing the adapter.
+        """
 
     def cleanup(self) -> None:  # pragma: no cover - default no-op
-        """Hook for cleaning up resources such as temporary files or servers."""
+        """
+        Clean up resources after execution.
+
+        This hook can be used to release resources such as temporary files
+        or running servers.
+
+        Returns:
+            None
+        """
 
     def supports_replay(self) -> bool:
         return True

src/agentunit/core/__init__.py

@@ -1,4 +1,6 @@
-"""Core components for AgentUnit."""
+"""
+Core components for AgentUnit.
+"""
 
 from agentunit.datasets.base import DatasetCase, DatasetSource
 from agentunit.reporting.results import ScenarioResult

src/agentunit/core/exceptions.py

@@ -1,15 +1,23 @@
-"""Custom exceptions for AgentUnit."""
+"""
+Custom exceptions for AgentUnit.
+"""
 
 from __future__ import annotations
 
 
 class AgentUnitError(Exception):
-    """Base class for AgentUnit exceptions."""
+    """
+    Base class for AgentUnit exceptions.
+    """
 
 
 class AdapterNotAvailableError(AgentUnitError):
-    """Raised when an adapter cannot be initialized due to missing dependencies."""
+    """
+    Raised when an adapter cannot be initialized due to missing dependencies.
+    """
 
 
 class ScenarioExecutionError(AgentUnitError):
-    """Raised when a scenario fails during execution."""
+    """
+    Raised when a scenario fails during execution.
+    """

src/agentunit/core/replay.py

@@ -1,4 +1,6 @@
-"""Replay utilities leveraging stored traces."""
+"""
+Replay utilities leveraging stored traces.
+"""
 
 from __future__ import annotations
 
@@ -8,7 +10,9 @@
 
 
 def load_traces(traces_dir: str | Path) -> list[TraceLog]:
-    """Load stored traces from disk for deterministic replay or analysis."""
+    """
+    Load stored traces from disk for deterministic replay or analysis.
+    """
 
     path = Path(traces_dir)
     logs: list[TraceLog] = []

src/agentunit/core/runner.py

@@ -1,4 +1,6 @@
-"""Scenario runner orchestration."""
+"""
+Scenario runner orchestration.
+"""
 
 from __future__ import annotations
 

src/agentunit/core/scenario.py

@@ -1,4 +1,6 @@
-"""Scenario definition API exposed to end users."""
+"""
+Scenario definition API exposed to end users.
+"""
 
 from __future__ import annotations
 
@@ -19,7 +21,9 @@
 
 @dataclass(slots=True)
 class Scenario:
-    """Defines a reproducible agent evaluation scenario."""
+    """
+    Defines a reproducible agent evaluation scenario.
+    """
 
     name: str
     adapter: BaseAdapter
@@ -75,7 +79,9 @@ def from_crewai(
         name: str | None = None,
         **options: object,
     ) -> Scenario:
-        """Create scenario from CrewAI crew."""
+        """
+        Create scenario from CrewAI crew.
+        """
         from agentunit.adapters.crewai import CrewAIAdapter
 
         adapter = CrewAIAdapter.from_crew(crew, **options)
@@ -91,7 +97,9 @@ def from_autogen(
         name: str | None = None,
         **options: object,
     ) -> Scenario:
-        """Create scenario from AutoGen orchestrator."""
+        """
+        Create scenario from AutoGen orchestrator.
+        """
         from agentunit.adapters.autogen import AutoGenAdapter
 
         adapter = AutoGenAdapter(orchestrator=orchestrator, **options)

src/agentunit/core/trace.py

@@ -1,4 +1,6 @@
-"""Tracing utilities shared between adapters and the runner."""
+"""
+Tracing utilities shared between adapters and the runner.
+"""
 
 from __future__ import annotations
 
@@ -11,7 +13,9 @@
 
 @dataclass(slots=True)
 class TraceEvent:
-    """Represents a single prompt, tool call, or response in an agent run."""
+    """
+    Represents a single prompt, tool call, or response in an agent run.
+    """
 
     type: str
     payload: dict[str, Any]
@@ -20,7 +24,9 @@ class TraceEvent:
 
 @dataclass(slots=True)
 class TraceLog:
-    """A collection of chronological events for a scenario iteration."""
+    """
+    A collection of chronological events for a scenario iteration.
+    """
 
     events: list[TraceEvent] = field(default_factory=list)
 

src/agentunit/py.typed

@@ -1,2 +0,0 @@
-# Marker file for PEP 561
-# This file indicates that the agentunit package supports type checking