diff --git a/docs/docs/07_loop_agents.md b/docs/docs/07_loop_agents.md
index 78b7fa3..bbd3bd2 100644
--- a/docs/docs/07_loop_agents.md
+++ b/docs/docs/07_loop_agents.md
@@ -182,7 +182,7 @@ def exit_loop(tool_context: ToolContext):
     Called by the refiner when critic approves the essay.
     """
     print(f"  [Exit Loop] Called by {tool_context.agent_name} - Essay approved!")
-    tool_context.actions.end_of_agent = True  # Signal to stop looping
+    tool_context.actions.escalate = True  # Signal to stop looping
     # Return a minimal valid content part so the backend always produces a valid LlmResponse
     return {"text": "Loop exited successfully. The agent has determined the task is complete."}
 
diff --git a/tutorial_implementation/tutorial07/essay_refiner/agent.py b/tutorial_implementation/tutorial07/essay_refiner/agent.py
index 2688774..3e03066 100644
--- a/tutorial_implementation/tutorial07/essay_refiner/agent.py
+++ b/tutorial_implementation/tutorial07/essay_refiner/agent.py
@@ -10,7 +10,7 @@ def exit_loop(tool_context: ToolContext):
     Called by the refiner when critic approves the essay.
     """
     print(f"  [Exit Loop] Called by {tool_context.agent_name} - Essay approved!")
-    tool_context.actions.end_of_agent = True  # Signal to stop looping
+    tool_context.actions.escalate = True  # Signal to stop looping
     # Return a minimal valid content part so the backend always produces a valid LlmResponse
     return {"text": "Loop exited successfully. The agent has determined the task is complete."}
 
@@ -126,4 +126,4 @@ def exit_loop(tool_context: ToolContext):
 )
 
 # MUST be named root_agent for ADK
-root_agent = essay_refinement_system
\ No newline at end of file
+root_agent = essay_refinement_system