diff --git a/pr_agent/git_providers/git_provider.py b/pr_agent/git_providers/git_provider.py
index 38b532ef..2b5e4e71 100644
--- a/pr_agent/git_providers/git_provider.py
+++ b/pr_agent/git_providers/git_provider.py
@@ -75,20 +75,45 @@ class GitProvider(ABC):
     def get_user_description(self) -> str:
         description = (self.get_pr_description_full() or "").strip()
         description_lowercase = description.lower()
+
         # if the existing description wasn't generated by the pr-agent, just return it as-is
         if not self._is_generated_by_pr_agent(description_lowercase):
             return description
-        # if the existing description was generated by the pr-agent, but it doesn't contain the user description,
+
+        # if the existing description was generated by the pr-agent, but it doesn't contain a user description,
         # return nothing (empty string) because it means there is no user description
         user_description_header = "## user description"
         if user_description_header not in description_lowercase:
             return ""
+
         # otherwise, extract the original user description from the existing pr-agent description and return it
-        user_description_start_position = description_lowercase.find(user_description_header) + len(user_description_header)
-        return description[user_description_start_position:].split("\n", 1)[-1].strip()
+        # user_description_start_position = description_lowercase.find(user_description_header) + len(user_description_header)
+        # return description[user_description_start_position:].split("\n", 1)[-1].strip()
+
+        # the 'user description' is in the beginning. extract and return it
+        possible_headers = self._possible_headers()
+        start_position = description_lowercase.find(user_description_header) + len(user_description_header)
+        end_position = len(description)
+        for header in possible_headers: # try to clip at the next header
+            if header != user_description_header and header in description_lowercase:
+                end_position = min(end_position, description_lowercase.find(header))
+        if end_position != len(description) and end_position > start_position:
+            original_user_description = description[start_position:end_position].strip()
+            if original_user_description.endswith("___"):
+                original_user_description = original_user_description[:-3].strip()
+        else:
+            original_user_description = description.split("___")[0].strip()
+            if original_user_description.lower().startswith(user_description_header):
+                original_user_description = original_user_description[len(user_description_header):].strip()
+
+        return original_user_description
+
+    def _possible_headers(self):
+        return ("## user description", "## pr type", "## pr description", "## pr labels", "## type", "## description",
+                "## labels", "### 🤖 generated by pr agent")
 
     def _is_generated_by_pr_agent(self, description_lowercase: str) -> bool:
-        possible_headers = ("## pr type", "## pr description", "## pr labels", "## type", "## description", "## labels", "### 🤖 generated by pr agent")
+        possible_headers = self._possible_headers()
         return any(description_lowercase.startswith(header) for header in possible_headers)
 
     @abstractmethod
diff --git a/pr_agent/tools/pr_description.py b/pr_agent/tools/pr_description.py
index 362849d5..858b8921 100644
--- a/pr_agent/tools/pr_description.py
+++ b/pr_agent/tools/pr_description.py
@@ -185,7 +185,12 @@ class PRDescription:
         # Load the AI prediction data into a dictionary
         self.data = load_yaml(self.prediction.strip())
 
+        if get_settings().pr_description.add_original_user_description and self.user_description:
+            self.data["User Description"] = self.user_description
+
         # re-order keys
+        if 'User Description' in self.data:
+            self.data['User Description'] = self.data.pop('User Description')
         if 'title' in self.data:
             self.data['title'] = self.data.pop('title')
         if 'type' in self.data:
@@ -197,8 +202,7 @@ class PRDescription:
         if 'pr_files' in self.data:
             self.data['pr_files'] = self.data.pop('pr_files')
 
-        if get_settings().pr_description.add_original_user_description and self.user_description:
-            self.data["User Description"] = self.user_description
+
 
 
     def _prepare_labels(self) -> List[str]: