diff --git a/src/interface.jl b/src/interface.jl
index 304a00f..c662a71 100755
--- a/src/interface.jl
+++ b/src/interface.jl
@@ -224,7 +224,6 @@ function agentReflex(
       """,
   ),
   thinkingFormat::Dict=Dict(
-    :no_thinking=> "",
     :react=> 
       """Use the following format:
       Question: the input question your user is asking and you must answer
@@ -460,8 +459,7 @@ function generatePrompt_mistral_openorca(a::T, usermsg::String,
   return prompt
 end
 
-
-function genPrompt_mistral_openorca_planning(a::agentReflex, usermsg::String) #WORKING
+function genPrompt_mistral_openorca(a::agentReflex, usermsg::String) #WORKING
   """
     general prompt format:
 
@@ -480,8 +478,74 @@ function genPrompt_mistral_openorca_planning(a::agentReflex, usermsg::String) #W
     "
 
     Note:
-      {context} = {earlierConversation} + {current status} + 
-                  {shortterm memory} + {longterm memory}
+      {context} = 
+        "
+        {earlierConversation}
+        {current status}
+        {shortterm memory}
+        {longterm memory}
+        "
+  """
+
+  prompt = 
+  """
+    <|im_start|>system
+    {role}
+    {tools}
+    {thinkingFormat}
+    <|im_end|>
+    {context}
+    <|im_start|>user
+    {usermsg}
+    <|im_end|>
+    <|im_start|>assistant
+
+  """
+  prompt = replace(prompt, "{role}" => a.roles[a.role])
+  prompt = replace(prompt, "{thinkingFormat}" => "")
+
+  context = 
+  """
+  {earlierConversation}
+  {current status}
+  {longterm memory}
+  """
+  prompt = replace(prompt, "{earlierConversation}" => "My earlier talk with the user:\n$(a.earlierConversation)")
+  prompt = replace(prompt, "{current status}" => "")
+  prompt = replace(prompt, "{longterm memory}" => "")
+
+  prompt *= "<|im_start|>user\nQuestion: " * usermsg * "\n<|im_end|>\n"
+  prompt *= "<|im_start|>assistant\n"
+  
+  return prompt
+end
+
+function genPrompt_planning_mistral_openorca(a::agentReflex, usermsg::String) #WORKING
+  """
+    general prompt format:
+
+    "
+    <|im_start|>system
+    {role}
+    {tools}
+    {thinkingFormat}
+    <|im_end|>
+    {context}
+    <|im_start|>user
+    {usermsg}
+    <|im_end|>
+    <|im_start|>assistant
+
+    "
+
+    Note:
+      {context} = 
+        "
+        {earlierConversation}
+        {current status}
+        {shortterm memory}
+        {longterm memory}
+        "
   """
 
   prompt = 
@@ -731,7 +795,7 @@ function conversation(a::agentReflex, usermsg::String; thinkingroundlimit::Int=3
   if a.thinkingMode == :no_thinking
     a.earlierConversation = conversationSummary(a)  #TODO should be long conversation before use summary because it leaves out details
     _ = addNewMessage(a, "user", usermsg)
-    prompt = generatePrompt_mistral_openorca(a, usermsg)
+    prompt = genPrompt_mistral_openorca(a, usermsg)  #TODO rewrite this function
     @show prompt
     respond = sendReceivePrompt(a, prompt)
     respond = split(respond, "<|im_end|>")[1]
@@ -763,7 +827,7 @@ function work(a::agentReflex, usermsg::String)
     @show a.thinkinground
     toolname = nothing
     toolinput = nothing
-    plan = generatePrompt_planning(a, prompt)
+    plan = genPrompt_planning_mistral_openorca(a, prompt)
     @show plan
     # for 
     #   # execute