diff --git a/src/crewai/agents/agent_state.py b/src/crewai/agents/agent_state.py
index f83fa86db..b75d81e69 100644
--- a/src/crewai/agents/agent_state.py
+++ b/src/crewai/agents/agent_state.py
@@ -23,24 +23,14 @@ class AgentState(BaseModel):
     """
 
     # Core fields
-    completed: bool = Field(
-        default=False,
-        description="Whether the current task is finished"
-    )
-
     original_plan: List[str] = Field(
         default_factory=list,
         description="The initial plan from first reasoning pass. Never overwrite unless user requests complete replan"
     )
 
-    last_plan: List[str] = Field(
-        default_factory=list,
-        description="The most recent plan (original or mid-execution update)"
-    )
-
     acceptance_criteria: List[str] = Field(
         default_factory=list,
-        description="Concrete goals to satisfy before marking completed=true"
+        description="Concrete goals to satisfy for task completion"
     )
 
     scratchpad: Dict[str, Any] = Field(
@@ -74,16 +64,10 @@ class AgentState(BaseModel):
         description="Number of execution steps completed"
     )
 
-    def update_last_plan(self, new_plan: List[str]) -> None:
-        """Update the last plan and timestamp."""
-        self.last_plan = new_plan
-        self.last_updated = datetime.now()
-
     def set_original_plan(self, plan: List[str]) -> None:
         """Set the original plan (only if not already set)."""
         if not self.original_plan:
             self.original_plan = plan
-            self.last_plan = plan
             self.last_updated = datetime.now()
 
     def add_to_scratchpad(self, key: str, value: Any) -> None:
@@ -132,16 +116,9 @@ class AgentState(BaseModel):
         self.steps_completed += 1
         self.last_updated = datetime.now()
 
-    def mark_completed(self) -> None:
-        """Mark the task as completed."""
-        self.completed = True
-        self.last_updated = datetime.now()
-
     def reset(self, task_id: Optional[str] = None) -> None:
         """Reset state for a new task."""
-        self.completed = False
         self.original_plan = []
-        self.last_plan = []
         self.acceptance_criteria = []
         self.scratchpad = {}
         self.tool_usage_history = []
@@ -154,16 +131,15 @@ class AgentState(BaseModel):
         """Generate a concise string representation for LLM context."""
         context = f"Current State (Step {self.steps_completed}):\n"
         context += f"- Task ID: {self.task_id}\n"
-        context += f"- Completed: {self.completed}\n"
 
         if self.acceptance_criteria:
             context += "- Acceptance Criteria:\n"
             for criterion in self.acceptance_criteria:
                 context += f"  • {criterion}\n"
 
-        if self.last_plan:
-            context += "- Current Plan:\n"
-            for i, step in enumerate(self.last_plan, 1):
+        if self.original_plan:
+            context += "- Plan:\n"
+            for i, step in enumerate(self.original_plan, 1):
                 context += f"  {i}. {step}\n"
 
         if self.tool_usage_history:
diff --git a/src/crewai/agents/crew_agent_executor.py b/src/crewai/agents/crew_agent_executor.py
index 16aedc6fe..7492984d8 100644
--- a/src/crewai/agents/crew_agent_executor.py
+++ b/src/crewai/agents/crew_agent_executor.py
@@ -138,9 +138,6 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
         if self.ask_for_human_input:
             formatted_answer = self._handle_human_feedback(formatted_answer)
 
-        # Mark task as completed in agent state
-        self.agent_state.mark_completed()
-
         self._create_short_term_memory(formatted_answer)
         self._create_long_term_memory(formatted_answer)
         self._create_external_memory(formatted_answer)
@@ -312,6 +309,7 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
                                 f"Agent State:\n"
                                 f"Raw: {self.agent_state.model_dump_json()}\n"
                                 f"[AGENT STATE] Step {self.agent_state.steps_completed}:\n"
+                                f"Original Plan: {getattr(self.agent_state, 'original_plan', None)}\n"
                                 f"Tool: {formatted_answer.tool}\n"
                                 f"Scratchpad: {self.agent_state.scratchpad}\n"
                                 f"Tool History: {len(self.agent_state.tool_usage_history)} entries\n"
@@ -749,12 +747,8 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
                 iteration_messages=self.messages,
             )
 
-            # Update agent state with new plan if available
+            # Update acceptance criteria if they changed from the reasoning output
             if reasoning_output.plan.structured_plan:
-                self.agent_state.update_last_plan(
-                    reasoning_output.plan.structured_plan.steps
-                )
-                # Update acceptance criteria if they changed
                 if reasoning_output.plan.structured_plan.acceptance_criteria:
                     self.agent_state.acceptance_criteria = (
                         reasoning_output.plan.structured_plan.acceptance_criteria