diff --git a/tests/agent/guardrails/actions/test_block_action.py b/tests/agent/guardrails/actions/test_block_action.py
index d6b3f1e2..988fc964 100644
--- a/tests/agent/guardrails/actions/test_block_action.py
+++ b/tests/agent/guardrails/actions/test_block_action.py
@@ -35,3 +35,69 @@ async def test_node_name_and_exception_pre_llm(self):
 
         # The exception string is the provided reason
         assert str(excinfo.value) == "Sensitive data detected"
+
+    @pytest.mark.asyncio
+    async def test_node_name_and_exception_post_llm(self):
+        """PostExecution + LLM: name is sanitized and node raises correct exception."""
+        action = BlockAction(reason="Invalid output detected")
+        guardrail = MagicMock()
+        guardrail.name = "Output Guardrail v2"
+
+        node_name, node = action.action_node(
+            guardrail=guardrail,
+            scope=GuardrailScope.LLM,
+            execution_stage=ExecutionStage.POST_EXECUTION,
+            guarded_component_name="guarded_node_name",
+        )
+
+        assert node_name == "llm_post_execution_output_guardrail_v2_block"
+
+        with pytest.raises(AgentTerminationException) as excinfo:
+            await node(AgentGuardrailsGraphState(messages=[]))
+
+        # The exception string is the provided reason
+        assert str(excinfo.value) == "Invalid output detected"
+
+    @pytest.mark.asyncio
+    async def test_node_name_and_exception_pre_tool(self):
+        """PreExecution + TOOL: name is sanitized and node raises correct exception."""
+        action = BlockAction(reason="Tool input validation failed")
+        guardrail = MagicMock()
+        guardrail.name = "Tool-Safety@Check"
+
+        node_name, node = action.action_node(
+            guardrail=guardrail,
+            scope=GuardrailScope.TOOL,
+            execution_stage=ExecutionStage.PRE_EXECUTION,
+            guarded_component_name="test_tool",
+        )
+
+        assert node_name == "tool_pre_execution_tool_safety_check_block"
+
+        with pytest.raises(AgentTerminationException) as excinfo:
+            await node(AgentGuardrailsGraphState(messages=[]))
+
+        # The exception string is the provided reason
+        assert str(excinfo.value) == "Tool input validation failed"
+
+    @pytest.mark.asyncio
+    async def test_node_name_and_exception_post_tool(self):
+        """PostExecution + TOOL: name is sanitized and node raises correct exception."""
+        action = BlockAction(reason="Tool output validation failed")
+        guardrail = MagicMock()
+        guardrail.name = "Output-Validator#2024"
+
+        node_name, node = action.action_node(
+            guardrail=guardrail,
+            scope=GuardrailScope.TOOL,
+            execution_stage=ExecutionStage.POST_EXECUTION,
+            guarded_component_name="test_tool",
+        )
+
+        assert node_name == "tool_post_execution_output_validator_2024_block"
+
+        with pytest.raises(AgentTerminationException) as excinfo:
+            await node(AgentGuardrailsGraphState(messages=[]))
+
+        # The exception string is the provided reason
+        assert str(excinfo.value) == "Tool output validation failed"
diff --git a/tests/agent/guardrails/actions/test_log_action.py b/tests/agent/guardrails/actions/test_log_action.py
index cf172ff3..48117a69 100644
--- a/tests/agent/guardrails/actions/test_log_action.py
+++ b/tests/agent/guardrails/actions/test_log_action.py
@@ -78,3 +78,188 @@ async def test_default_message_includes_context(
             == "Guardrail [My Guardrail] validation failed for [TOOL] [POST_EXECUTION] with the following reason: bad input"
             for rec in caplog.records
         )
+
+    @pytest.mark.asyncio
+    async def test_node_name_and_exception_post_llm(
+        self, caplog: pytest.LogCaptureFixture
+    ) -> None:
+        """PostExecution + LLM: name is sanitized and default message is logged."""
+        action = LogAction(message=None, level=logging.INFO)
+        guardrail = MagicMock()
+        guardrail.name = "Test Guardrail"
+
+        node_name, node = action.action_node(
+            guardrail=guardrail,
+            scope=GuardrailScope.LLM,
+            execution_stage=ExecutionStage.POST_EXECUTION,
+            guarded_component_name="guarded_node_name",
+        )
+
+        # Verify node name format
+        assert node_name == "llm_post_execution_test_guardrail_log"
+        assert isinstance(node_name, str)
+        assert node_name.endswith("_log")
+        assert "llm" in node_name
+        assert "post_execution" in node_name
+
+        # Verify node is callable
+        assert callable(node)
+
+        # Verify node returns empty dict
+        with caplog.at_level(logging.INFO):
+            await node(
+                AgentGuardrailsGraphState(
+                    messages=[], guardrail_validation_result="validation error"
+                )
+            )
+
+        # Verify log record properties
+        log_record = caplog.records[0]
+        assert log_record.levelno == logging.INFO
+
+        # Verify default message includes all context
+        assert (
+            "Guardrail [Test Guardrail] validation failed for [LLM] [POST_EXECUTION]"
+            in log_record.message
+        )
+        assert "validation error" in log_record.message
+        assert (
+            log_record.message
+            == "Guardrail [Test Guardrail] validation failed for [LLM] [POST_EXECUTION] with the following reason: validation error"
+        )
+
+    @pytest.mark.asyncio
+    async def test_node_name_and_exception_pre_tool(
+        self, caplog: pytest.LogCaptureFixture
+    ) -> None:
+        """PreExecution + TOOL: name is sanitized and default message is logged."""
+        action = LogAction(message=None, level=logging.WARNING)
+        guardrail = MagicMock()
+        guardrail.name = "Tool Guardrail v2"
+
+        node_name, node = action.action_node(
+            guardrail=guardrail,
+            scope=GuardrailScope.TOOL,
+            execution_stage=ExecutionStage.PRE_EXECUTION,
+            guarded_component_name="test_tool",
+        )
+
+        # Verify node name format
+        assert node_name == "tool_pre_execution_tool_guardrail_v2_log"
+        assert isinstance(node_name, str)
+        assert node_name.endswith("_log")
+        assert "tool" in node_name
+        assert "pre_execution" in node_name
+
+        # Verify node returns empty dict
+        with caplog.at_level(logging.WARNING):
+            await node(
+                AgentGuardrailsGraphState(
+                    messages=[], guardrail_validation_result="invalid tool args"
+                )
+            )
+
+        # Verify log record properties
+        log_record = caplog.records[0]
+        assert log_record.levelno == logging.WARNING
+
+        # Verify default message includes all context
+        assert (
+            "Guardrail [Tool Guardrail v2] validation failed for [TOOL] [PRE_EXECUTION]"
+            in log_record.message
+        )
+        assert "invalid tool args" in log_record.message
+        assert (
+            log_record.message
+            == "Guardrail [Tool Guardrail v2] validation failed for [TOOL] [PRE_EXECUTION] with the following reason: invalid tool args"
+        )
+
+    @pytest.mark.asyncio
+    async def test_node_name_and_exception_post_tool(
+        self, caplog: pytest.LogCaptureFixture
+    ) -> None:
+        """PostExecution + TOOL: name is sanitized and custom message is logged."""
+        action = LogAction(message="Tool execution failed", level=logging.ERROR)
+        guardrail = MagicMock()
+        guardrail.name = "Special-Tool@Guardrail"
+
+        node_name, node = action.action_node(
+            guardrail=guardrail,
+            scope=GuardrailScope.TOOL,
+            execution_stage=ExecutionStage.POST_EXECUTION,
+            guarded_component_name="test_tool",
+        )
+
+        # Verify node name format (special characters are sanitized)
+        assert node_name == "tool_post_execution_special_tool_guardrail_log"
+        assert isinstance(node_name, str)
+        assert node_name.endswith("_log")
+        assert "tool" in node_name
+        assert "post_execution" in node_name
+
+        # Verify node returns empty dict
+        with caplog.at_level(logging.ERROR):
+            await node(
+                AgentGuardrailsGraphState(
+                    messages=[], guardrail_validation_result="tool error"
+                )
+            )
+
+        # Verify log record properties
+        log_record = caplog.records[0]
+        assert log_record.levelno == logging.ERROR
+
+        # Verify custom message was logged (not default message)
+        assert log_record.message == "Tool execution failed"
+        assert (
+            "Guardrail" not in log_record.message
+        )  # Custom message doesn't include guardrail context
+        assert (
+            "validation failed" not in log_record.message
+        )  # Custom message doesn't include default format
+
+    @pytest.mark.asyncio
+    async def test_node_name_and_exception_post_tool_warning(
+        self, caplog: pytest.LogCaptureFixture
+    ) -> None:
+        """PostExecution + TOOL: name is sanitized and default message is logged at WARNING level."""
+        action = LogAction(message=None, level=logging.WARNING)
+        guardrail = MagicMock()
+        guardrail.name = "Post Tool Guardrail"
+
+        node_name, node = action.action_node(
+            guardrail=guardrail,
+            scope=GuardrailScope.TOOL,
+            execution_stage=ExecutionStage.POST_EXECUTION,
+            guarded_component_name="test_tool",
+        )
+
+        # Verify node name format
+        assert node_name == "tool_post_execution_post_tool_guardrail_log"
+        assert isinstance(node_name, str)
+        assert node_name.endswith("_log")
+        assert "tool" in node_name
+        assert "post_execution" in node_name
+
+        # Verify node returns empty dict
+        with caplog.at_level(logging.WARNING):
+            await node(
+                AgentGuardrailsGraphState(
+                    messages=[], guardrail_validation_result="post execution error"
+                )
+            )
+
+        # Verify log record properties
+        log_record = caplog.records[0]
+        assert log_record.levelno == logging.WARNING
+
+        # Verify default message includes all context
+        assert (
+            "Guardrail [Post Tool Guardrail] validation failed for [TOOL] [POST_EXECUTION]"
+            in log_record.message
+        )
+        assert "post execution error" in log_record.message
+        assert (
+            log_record.message
+            == "Guardrail [Post Tool Guardrail] validation failed for [TOOL] [POST_EXECUTION] with the following reason: post execution error"
+        )