fix(tools): make crew-scoped policy real and close the last raise leak

Two findings, and the first was a documented feature that never worked.

`resolve_tool_failure_policy` consulted a crew, and the docs advertised
crew as a scope, but `Crew` had no `tool_failure_policy` field at all --
and even with one it was unreachable, because `BaseAgent` defaulted the
policy to `WARN` rather than `None`, so resolution always stopped at the
agent. Crew-level configuration was silently ignored.

Fixed by making "inherit" the default everywhere instead of baking `warn`
into one layer: `Crew` gains the field, and `BaseAgent`/`LiteAgent`
default to `None` like `Task` and `BaseTool` already did. The resolver
owns the single fallback, so the chain is genuinely
tool > task > agent > crew > warn and the effective default with nothing
configured is still `warn`. Reading `agent.tool_failure_policy` now
returns `None` (meaning "inherit") rather than `WARN`.

The other: `StepExecutor` re-raised `ToolExecutionFailedError` from its
outer handler, but the nested handler around the native-to-text tooling
fallback still caught it and returned `StepResult(success=False)`. An
agent whose LLM lacked native tool calling would therefore not abort
under `raise`. That is the third distinct place this exception was being
downgraded; it now re-raises there too.

Testing: 8 further tests, 60 total, including the full precedence chain
walked one level at a time and crew-scoped `raise`/`ignore` driven
end-to-end through `kickoff()` rather than only through the resolver --
the gap that let the original crew bug pass review. Full suite matches
baseline exactly at 377 pre-existing failures; mypy clean on every
changed file.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01ETacm2dMASfpMAYUiDu5YG

docs/edge/en/concepts/tools.mdx

@@ -388,7 +388,7 @@ tool's `max_usage_count` is spent, or when the agent calls a tool that does not
 | `raise` | Records and emits, then aborts with `ToolExecutionFailedError`. |
 
 ```python Code
-from crewai import Agent, Task
+from crewai import Agent, Crew, Task
 from crewai.tools.tool_failure import ToolFailurePolicy
 
 agent = Agent(
@@ -406,10 +406,18 @@ task = Task(
     agent=agent,
     tool_failure_policy=ToolFailurePolicy.RAISE,
 )
+
+# Or set a baseline once for every agent in the crew.
+crew = Crew(
+    agents=[agent],
+    tasks=[task],
+    tool_failure_policy=ToolFailurePolicy.WARN,
+)
 ```
 
-The most specific setting wins: tool, then task, then agent, then crew, then the
-`warn` default.
+The most specific setting wins: **tool → task → agent → crew → `warn`**. Every
+level defaults to `None`, meaning "inherit from the next one out", so the
+effective default with nothing configured anywhere is `warn`.
 
 ### Inspecting Failures
 

lib/crewai/src/crewai/agents/agent_builder/base_agent.py

@@ -300,8 +300,8 @@ class BaseAgent(BaseModel, ABC, metaclass=AgentMeta):
     max_iter: int = Field(
         default=25, description="Maximum iterations for an agent to execute a task"
     )
-    tool_failure_policy: ToolFailurePolicy = Field(
-        default=ToolFailurePolicy.WARN,
+    tool_failure_policy: ToolFailurePolicy | None = Field(
+        default=None,
         description=(
             "How to react when a tool runs to completion but reports that it "
             "failed (an upstream API rejecting the request, an MCP server "
@@ -309,7 +309,8 @@ class BaseAgent(BaseModel, ABC, metaclass=AgentMeta):
             "'ignore' restores pre-1.16 behavior and records nothing; 'warn' "
             "records the failure, emits ToolFailureDetectedEvent and keeps "
             "going; 'raise' additionally aborts with ToolExecutionFailedError. "
-            "A Task or a tool may override this for a narrower scope."
+            "None inherits from the crew, falling back to 'warn'. A task or a "
+            "tool may override this for a narrower scope."
         ),
     )
     agent_executor: Annotated[

lib/crewai/src/crewai/agents/step_executor.py

@@ -224,6 +224,12 @@ class StepExecutor:
                         tool_calls_made=tool_calls_made,
                         execution_time=elapsed,
                     )
+                except ToolExecutionFailedError:
+                    # Same reason as the outer handler: a deliberate stop must
+                    # not be downgraded into StepResult(success=False), even
+                    # when reached through the text-tooling fallback.
+                    raise
+
                 except Exception as fallback_error:
                     e = fallback_error
 

lib/crewai/src/crewai/crew.py

@@ -116,6 +116,7 @@ from crewai.tasks.task_output import TaskOutput
 from crewai.tools.agent_tools.agent_tools import AgentTools
 from crewai.tools.agent_tools.read_file_tool import ReadFileTool
 from crewai.tools.base_tool import BaseTool
+from crewai.tools.tool_failure import ToolFailurePolicy
 from crewai.types.callback import SerializableCallable
 from crewai.types.streaming import CrewStreamingOutput
 from crewai.types.usage_metrics import UsageMetrics
@@ -231,6 +232,15 @@ class Crew(FlowTrackable, BaseModel):
             "unless they set a cache_function that prevents caching."
         ),
     )
+    tool_failure_policy: ToolFailurePolicy | None = Field(
+        default=None,
+        description=(
+            "Baseline reaction for every agent in this crew when a tool runs "
+            "to completion but reports that it failed. Leave None for the "
+            "'warn' default. An agent, task, or tool may override it for a "
+            "narrower scope."
+        ),
+    )
     tasks: list[Task] = Field(default_factory=list)
     agents: Annotated[
         list[BaseAgent],

lib/crewai/src/crewai/lite_agent.py

@@ -228,11 +228,12 @@ class LiteAgent(FlowTrackable, BaseModel):
     max_iterations: int = Field(
         default=15, description="Maximum number of iterations for tool usage"
     )
-    tool_failure_policy: ToolFailurePolicy = Field(
-        default=ToolFailurePolicy.WARN,
+    tool_failure_policy: ToolFailurePolicy | None = Field(
+        default=None,
         description=(
             "How to react when a tool runs to completion but reports that it "
-            "failed. See BaseAgent.tool_failure_policy."
+            "failed. None falls back to 'warn'. See "
+            "BaseAgent.tool_failure_policy."
         ),
     )
     max_execution_time: int | None = Field(

lib/crewai/tests/tools/test_tool_failure.py

@@ -173,6 +173,62 @@ class TestPolicyResolution:
         resolved = resolve_tool_failure_policy(agent=agent, task=task)
         assert resolved is ToolFailurePolicy.IGNORE
 
+    def test_crew_policy_used_when_agent_inherits(self) -> None:
+        from crewai import Crew
+
+        agent = Agent(role="r", goal="g", backstory="b")
+        crew = Crew(
+            agents=[agent], tasks=[], tool_failure_policy=ToolFailurePolicy.RAISE
+        )
+        resolved = resolve_tool_failure_policy(agent=agent, crew=crew)
+        assert resolved is ToolFailurePolicy.RAISE
+
+    def test_agent_overrides_crew(self) -> None:
+        from crewai import Crew
+
+        agent = Agent(
+            role="r",
+            goal="g",
+            backstory="b",
+            tool_failure_policy=ToolFailurePolicy.IGNORE,
+        )
+        crew = Crew(
+            agents=[agent], tasks=[], tool_failure_policy=ToolFailurePolicy.RAISE
+        )
+        resolved = resolve_tool_failure_policy(agent=agent, crew=crew)
+        assert resolved is ToolFailurePolicy.IGNORE
+
+    def test_full_precedence_chain(self) -> None:
+        """tool > task > agent > crew > warn."""
+        from crewai import Crew
+
+        class ScopedTool(SlackTool):
+            tool_failure_policy: ToolFailurePolicy | None = None
+
+        tool = ScopedTool()
+        agent = Agent(role="r", goal="g", backstory="b")
+        task = Task(description="d", expected_output="e")
+        crew = Crew(agents=[agent], tasks=[])
+
+        def resolved() -> ToolFailurePolicy:
+            return resolve_tool_failure_policy(
+                tool=tool, agent=agent, task=task, crew=crew
+            )
+
+        assert resolved() is ToolFailurePolicy.WARN
+
+        crew.tool_failure_policy = ToolFailurePolicy.IGNORE
+        assert resolved() is ToolFailurePolicy.IGNORE
+
+        agent.tool_failure_policy = ToolFailurePolicy.WARN
+        assert resolved() is ToolFailurePolicy.WARN
+
+        task.tool_failure_policy = ToolFailurePolicy.RAISE
+        assert resolved() is ToolFailurePolicy.RAISE
+
+        tool.tool_failure_policy = ToolFailurePolicy.IGNORE
+        assert resolved() is ToolFailurePolicy.IGNORE
+
     def test_invalid_policy_is_ignored_rather_than_raising(self) -> None:
         """A bad policy value must never take down a tool call."""
 
@@ -208,14 +264,28 @@ class TestPolicyResolution:
         assert resolved is ToolFailurePolicy.RAISE
 
 
-class TestAgentDefault:
-    def test_agent_defaults_to_warn(self) -> None:
-        agent = Agent(role="r", goal="g", backstory="b")
-        assert agent.tool_failure_policy is ToolFailurePolicy.WARN
+class TestDefaults:
+    """Every scope defaults to None ('inherit'); the resolver owns 'warn'."""
 
-    def test_task_policy_defaults_to_none_so_it_inherits(self) -> None:
+    def test_agent_defaults_to_inherit(self) -> None:
+        assert Agent(role="r", goal="g", backstory="b").tool_failure_policy is None
+
+    def test_task_defaults_to_inherit(self) -> None:
         assert Task(description="d", expected_output="e").tool_failure_policy is None
 
+    def test_crew_defaults_to_inherit(self) -> None:
+        from crewai import Crew
+
+        agent = Agent(role="r", goal="g", backstory="b")
+        assert Crew(agents=[agent], tasks=[]).tool_failure_policy is None
+
+    def test_tool_defaults_to_inherit(self) -> None:
+        assert SlackTool().tool_failure_policy is None
+
+    def test_effective_default_is_warn(self) -> None:
+        agent = Agent(role="r", goal="g", backstory="b")
+        assert resolve_tool_failure_policy(agent=agent) is ToolFailurePolicy.WARN
+
 
 class TestEndToEndPolicies:
     def test_warn_records_and_emits_without_stopping(self) -> None:
@@ -598,6 +668,41 @@ class TestRaisePolicySurvivesEveryWrapper:
             Crew(agents=[agent], tasks=[task]).kickoff()
         assert agent._times_executed == 0, "the abort must not trigger retries"
 
+    def test_crew_policy_aborts_end_to_end(self) -> None:
+        """Crew scope must actually reach the executor, not just the resolver."""
+        agent = Agent(
+            role="Slack Messenger",
+            goal="post a message",
+            backstory="b",
+            llm=ScriptedLLM(_slack_steps()),
+            tools=[SlackTool()],
+        )
+        task = Task(description="post to slack", expected_output="c", agent=agent)
+        crew = Crew(
+            agents=[agent],
+            tasks=[task],
+            tool_failure_policy=ToolFailurePolicy.RAISE,
+        )
+
+        with pytest.raises(ToolExecutionFailedError):
+            crew.kickoff()
+
+    def test_crew_ignore_suppresses_recording_end_to_end(self) -> None:
+        agent = Agent(
+            role="Slack Messenger",
+            goal="post a message",
+            backstory="b",
+            llm=ScriptedLLM(_slack_steps()),
+            tools=[SlackTool()],
+        )
+        task = Task(description="post to slack", expected_output="c", agent=agent)
+        result = Crew(
+            agents=[agent],
+            tasks=[task],
+            tool_failure_policy=ToolFailurePolicy.IGNORE,
+        ).kickoff()
+        assert not result.has_tool_failures
+
     def test_passthrough_tuple_includes_the_error(self) -> None:
         from crewai.agent.core import _passthrough_exceptions