Fix type annotation mismatch in Task.context field

src/crewai/crew.py

@@ -1034,7 +1034,7 @@ def _get_context(self, task: Task, task_outputs: List[TaskOutput]) -> str:
         context = (
             aggregate_raw_outputs_from_task_outputs(task_outputs)
             if task.context is NOT_SPECIFIED
-            else aggregate_raw_outputs_from_tasks(task.context)
+            else aggregate_raw_outputs_from_tasks(cast(List["Task"], task.context))
         )
         return context
 

src/crewai/task.py

@@ -26,6 +26,7 @@
 from pydantic import (
     UUID4,
     BaseModel,
+    ConfigDict,
     Field,
     PrivateAttr,
     field_validator,
@@ -39,7 +40,7 @@
 from crewai.tasks.task_output import TaskOutput
 from crewai.tools.base_tool import BaseTool
 from crewai.utilities.config import process_config
-from crewai.utilities.constants import NOT_SPECIFIED
+from crewai.utilities.constants import NOT_SPECIFIED, _NotSpecified
 from crewai.utilities.guardrail import process_guardrail, GuardrailResult
 from crewai.utilities.converter import Converter, convert_to_model
 from crewai.utilities.events import (
@@ -78,6 +79,8 @@ class Task(BaseModel):
     used_tools: int = 0
     tools_errors: int = 0
     delegations: int = 0
+
+    model_config = ConfigDict(arbitrary_types_allowed=True)
     i18n: I18N = I18N()
     name: Optional[str] = Field(default=None)
     prompt_context: Optional[str] = None
@@ -95,7 +98,7 @@ class Task(BaseModel):
     agent: Optional[BaseAgent] = Field(
         description="Agent responsible for execution the task.", default=None
     )
-    context: Optional[List["Task"]] = Field(
+    context: Union[List["Task"], _NotSpecified, None] = Field(
         description="Other tasks that will have their output used as context for this task.",
         default=NOT_SPECIFIED,
     )

tests/test_task_context_type_annotation.py

@@ -0,0 +1,103 @@
+"""
+Test for issue #3019: Type annotation for `context` in `Task` is `Optional[List["Task"]]`, but default is `NOT_SPECIFIED`
+
+This test reproduces the type annotation issue and verifies the fix.
+"""
+
+import pytest
+from typing import get_type_hints, get_origin, get_args
+from pydantic import ValidationError
+
+from crewai.task import Task
+from crewai.utilities.constants import NOT_SPECIFIED, _NotSpecified
+
+
+class TestTaskContextTypeAnnotation:
+    """Test cases for Task context field type annotation issue."""
+
+    def test_task_context_default_value_is_not_specified(self):
+        """Test that Task.context default value is NOT_SPECIFIED sentinel."""
+        task = Task(description="Test task", expected_output="Test output")
+        assert task.context is NOT_SPECIFIED
+        assert isinstance(task.context, _NotSpecified)
+
+    def test_task_context_can_be_set_to_none(self):
+        """Test that Task.context can be explicitly set to None."""
+        task = Task(description="Test task", expected_output="Test output", context=None)
+        assert task.context is None
+
+    def test_task_context_can_be_set_to_empty_list(self):
+        """Test that Task.context can be set to an empty list."""
+        task = Task(description="Test task", expected_output="Test output", context=[])
+        assert task.context == []
+        assert isinstance(task.context, list)
+
+    def test_task_context_can_be_set_to_task_list(self):
+        """Test that Task.context can be set to a list of tasks."""
+        task1 = Task(description="Task 1", expected_output="Output 1")
+        task2 = Task(description="Task 2", expected_output="Output 2")
+        task3 = Task(description="Task 3", expected_output="Output 3", context=[task1, task2])
+
+        assert task3.context == [task1, task2]
+        assert isinstance(task3.context, list)
+        assert len(task3.context) == 2
+
+    def test_task_context_type_annotation_includes_not_specified(self):
+        """Test that the type annotation for context includes _NotSpecified type."""
+        type_hints = get_type_hints(Task)
+        context_type = type_hints.get('context')
+
+        assert context_type is not None
+
+        origin = get_origin(context_type)
+        if origin is not None:
+            args = get_args(context_type)
+
+            assert any('_NotSpecified' in str(arg) or arg is _NotSpecified for arg in args), \
+                f"Type annotation should include _NotSpecified, got: {args}"
+
+    def test_task_context_distinguishes_not_passed_from_none(self):
+        """Test that NOT_SPECIFIED distinguishes between not passed and None."""
+        task_not_passed = Task(description="Test task", expected_output="Test output")
+
+        task_explicit_none = Task(description="Test task", expected_output="Test output", context=None)
+
+        task_empty_list = Task(description="Test task", expected_output="Test output", context=[])
+
+        assert task_not_passed.context is NOT_SPECIFIED
+        assert task_explicit_none.context is None
+        assert task_empty_list.context == []
+
+        assert task_not_passed.context is not task_explicit_none.context
+        assert task_not_passed.context != task_empty_list.context
+        assert task_explicit_none.context != task_empty_list.context
+
+    def test_task_context_usage_in_crew_logic(self):
+        """Test that the context field works correctly with crew logic."""
+        from crewai.utilities.constants import NOT_SPECIFIED
+
+        task_with_not_specified = Task(description="Task 1", expected_output="Output 1")
+        task_with_none = Task(description="Task 2", expected_output="Output 2", context=None)
+        task_with_empty_list = Task(description="Task 3", expected_output="Output 3", context=[])
+
+        assert task_with_not_specified.context is NOT_SPECIFIED
+        assert task_with_none.context is not NOT_SPECIFIED
+        assert task_with_empty_list.context is not NOT_SPECIFIED
+
+    def test_task_context_repr_shows_not_specified(self):
+        """Test that NOT_SPECIFIED has a proper string representation."""
+        task = Task(description="Test task", expected_output="Test output")
+        assert str(task.context) == "NOT_SPECIFIED"
+        assert repr(task.context) == "NOT_SPECIFIED"
+
+    def test_task_context_validation_accepts_valid_types(self):
+        """Test that Task validation accepts all valid context types."""
+        try:
+            Task(description="Test 1", expected_output="Output 1")
+            Task(description="Test 2", expected_output="Output 2", context=None)
+            Task(description="Test 3", expected_output="Output 3", context=[])
+
+            task1 = Task(description="Task 1", expected_output="Output 1")
+            Task(description="Test 4", expected_output="Output 4", context=[task1])
+        except ValidationError as e:
+            pytest.fail(f"Valid context types should not raise ValidationError: {e}")