diff --git a/loopgpt/agent.py b/loopgpt/agent.py
index f07d8f4..247b15d 100644
--- a/loopgpt/agent.py
+++ b/loopgpt/agent.py
@@ -59,6 +59,7 @@ def __init__(
         self.next_prompt = NEXT_PROMPT
         self.progress = []
         self.plan = []
+        self.criticism = None
         self.constraints = []
         self.state = AgentStates.START
 
@@ -125,6 +126,7 @@ def _get_compressed_history(self):
                     thoughts.pop("speak", None)
                     thoughts.pop("text", None)
                     thoughts.pop("plan", None)
+                    thoughts.pop("criticism", None)
                 entry["content"] = json.dumps(respd, indent=2)
                 hist[i] = entry
             except:
@@ -227,6 +229,9 @@ def chat(
                     self.plan = [plan]
                 if isinstance(plan, list):
                     self.plan = plan
+            criticism = resp.get("thoughts", {}).get("criticism")
+            if criticism:
+                self.criticism = criticism
         except:
             pass
         self.history.append({"role": "user", "content": message})
@@ -343,6 +348,7 @@ def clear_state(self):
         self.staging_tool = None
         self.staging_response = None
         self.tool_response = None
+        self.criticism = None
         self.progress = None
         self.state = AgentStates.START
         self.history.clear()
@@ -361,6 +367,8 @@ def header_prompt(self):
             prompt.append(self.constraints_prompt())
         if self.plan:
             prompt.append(self.plan_prompt())
+        if self.criticism:
+            prompt.append(self.criticism_prompt())
         if self.progress:
             prompt.append(self.progress_prompt())
         return "\n".join(prompt) + "\n"
@@ -379,6 +387,10 @@ def plan_prompt(self):
         plan = "\n".join(self.plan)
         return f"CURRENT PLAN:\n{plan}\n"
 
+    def criticism_prompt(self):
+        criticism = self.criticism
+        return f"CRITICISM: {criticism}"
+
     def goals_prompt(self):
         prompt = []
         prompt.append(f"GOALS:")
diff --git a/loopgpt/constants.py b/loopgpt/constants.py
index ca2d50a..780d34a 100644
--- a/loopgpt/constants.py
+++ b/loopgpt/constants.py
@@ -9,8 +9,9 @@
     "thoughts": {
         "text": "What do you want to say to the user?",
         "reasoning": "Why do you want to say this?",
-        "progress": "- A detailed list\n - of everything you have done so far",
+        "progress": "- A detailed list\n - of everything you have completed so far",
         "plan": "- short bulleted\n- list that conveys\n- long-term plan",
+        "criticism": "- A detailed list\n- of potential issues or inefficiencies\n- with your plan and reasoning",
         "speak": "thoughts summary to say to user",
     },
     "command": {"name": "next command in your plan", "args": {"arg name": "value"}},
diff --git a/loopgpt/loops/repl.py b/loopgpt/loops/repl.py
index d28cb7c..c48eb05 100644
--- a/loopgpt/loops/repl.py
+++ b/loopgpt/loops/repl.py
@@ -20,6 +20,7 @@
 LOOP_GPT = Fore.GREEN + "LoopGPT"
 REASONING = Fore.LIGHTBLUE_EX + "REASONING"
 PLAN = Fore.LIGHTYELLOW_EX + "PLAN"
+CRITICISM = Fore.RED + "CRITICISM"
 PROGRESS = Fore.LIGHTRED_EX + "PROGRESS"
 SPEAK = Fore.LIGHTGREEN_EX + "SPEAK"
 COMMAND = Fore.LIGHTMAGENTA_EX + "NEXT_COMMAND"
@@ -30,6 +31,7 @@
     "loopgpt": LOOP_GPT,
     "reasoning": REASONING,
     "plan": PLAN,
+    "criticism": CRITICISM,
     "progress": PROGRESS,
     "speak": SPEAK,
     "command": COMMAND,
@@ -132,6 +134,8 @@ def cli(agent, continuous=False):
                         if isinstance(thoughts["plan"], str)
                         else thoughts["plan"]
                     )
+                if "criticism" in thoughts:
+                    msgs["criticism"] = thoughts["criticism"]
                 if "progress" in thoughts:
                     msgs["progress"] = thoughts["progress"]
                 if "speak" in thoughts: