diff --git a/gitclone/enterprise b/gitclone/enterprise
new file mode 160000
index 0000000..38276c9
--- /dev/null
+++ b/gitclone/enterprise
@@ -0,0 +1 @@
+Subproject commit 38276c9a237a1779ce7a3e6f31102ff615691690
diff --git a/gitclone/odoo b/gitclone/odoo
new file mode 160000
index 0000000..8678e1c
--- /dev/null
+++ b/gitclone/odoo
@@ -0,0 +1 @@
+Subproject commit 8678e1c777b1faec23d4e7fd311c8ec96041f5b7
diff --git a/src/dataset_processor.py b/src/dataset_processor.py
index c5d3695..38b866c 100644
--- a/src/dataset_processor.py
+++ b/src/dataset_processor.py
@@ -133,8 +133,11 @@ class DatasetProcessor:
         Returns:
             List of code samples with metadata
         """
-        temp_dir = tempfile.mkdtemp()
-        self.temp_dirs.append(temp_dir)
+        # Create a persistent directory for cloned repositories
+        gitclone_dir = Path("./gitclone")
+        gitclone_dir.mkdir(exist_ok=True)
+        temp_dir = str(gitclone_dir)
+        # Note: We don't add this to temp_dirs since we want to keep it
 
         depth = 1
         branch = "18.0"
@@ -143,31 +146,31 @@ class DatasetProcessor:
             # Clone repository
             repo_name = repo_url.split('/')[-1].replace('.git', '')
             repo_path = os.path.join(temp_dir, repo_name)
+            if not os.path.exists(repo_path):
+                self.logger.info(f"Cloning {repo_url} to {repo_path}")
 
-            self.logger.info(f"Cloning {repo_url} to {repo_path}")
-            
-            # Use token for private repositories if provided
-            clone_url = repo_url
-            if github_token and "github.com" in repo_url:
-                # Handle SSH URLs
-                if repo_url.startswith("git@"):
-                    # SSH URL doesn't need token modification
-                    pass
-                else:
-                    # Add token to HTTPS URL
-                    if repo_url.startswith("https://"):
-                        clone_url = repo_url.replace("https://", f"https://{github_token}@")
-                    elif repo_url.startswith("http://"):
-                        clone_url = repo_url.replace("http://", f"http://{github_token}@")
+                # Use token for private repositories if provided
+                clone_url = repo_url
+                if github_token and "github.com" in repo_url:
+                    # Handle SSH URLs
+                    if repo_url.startswith("git@"):
+                        # SSH URL doesn't need token modification
+                        pass
                     else:
-                        # For URLs like "github.com/user/repo" or "user/repo"
-                        if repo_url.startswith("github.com/"):
-                            clone_url = f"https://{github_token}@{repo_url}"
+                        # Add token to HTTPS URL
+                        if repo_url.startswith("https://"):
+                            clone_url = repo_url.replace("https://", f"https://{github_token}@")
+                        elif repo_url.startswith("http://"):
+                            clone_url = repo_url.replace("http://", f"http://{github_token}@")
                         else:
-                            # Assume it's a GitHub path like "user/repo"
-                            clone_url = f"https://{github_token}@github.com/{repo_url}"
-            
-            repo = git.Repo.clone_from(clone_url, repo_path, depth=depth, branch=branch)
+                            # For URLs like "github.com/user/repo" or "user/repo"
+                            if repo_url.startswith("github.com/"):
+                                clone_url = f"https://{github_token}@{repo_url}"
+                            else:
+                                # Assume it's a GitHub path like "user/repo"
+                                clone_url = f"https://{github_token}@github.com/{repo_url}"
+                
+                repo = git.Repo.clone_from(clone_url, repo_path, depth=depth, branch=branch)
 
             # Extract code samples
             code_samples = self._extract_code_samples(repo_path, config)
@@ -175,8 +178,9 @@ class DatasetProcessor:
             return code_samples
 
         finally:
-            # Cleanup
-            shutil.rmtree(temp_dir, ignore_errors=True)
+            # Cleanup temporary directories, but keep gitclone folder
+            if temp_dir != "./gitclone":
+                shutil.rmtree(temp_dir, ignore_errors=True)
 
     def _extract_code_samples(self, repo_path: str, config: AppConfig) -> List[Dict]:
         """