Ensuring restricted access to the /learn index directory (#887)

* Add a test ensuring correct mode of index dir

* Ensure dir exists with correct mode

packages/jupyter-ai/jupyter_ai/chat_handlers/learn.py

@@ -19,6 +19,7 @@
     IndexMetadata,
 )
 from jupyter_core.paths import jupyter_data_dir
+from jupyter_core.utils import ensure_dir_exists
 from langchain.schema import BaseRetriever, Document
 from langchain.text_splitter import (
     LatexTextSplitter,
@@ -97,11 +98,12 @@ def __init__(self, *args, **kwargs):
         self.metadata = IndexMetadata(dirs=[])
         self.prev_em_id = None
 
-        if not os.path.exists(INDEX_SAVE_DIR):
-            os.makedirs(INDEX_SAVE_DIR)
-
+        self._ensure_dirs()
         self._load()
 
+    def _ensure_dirs(self):
+        ensure_dir_exists(INDEX_SAVE_DIR, mode=0o700)
+
     def _load(self):
         """Loads the vector store."""
         if self.index is not None:

packages/jupyter-ai/jupyter_ai/tests/test_handlers.py

@@ -1,3 +1,24 @@
+import os
+import stat
+from unittest import mock
+
+from jupyter_ai.chat_handlers import learn
+
+
+class MockLearnHandler(learn.LearnChatHandler):
+    def __init__(self):
+        pass
+
+
+def test_learn_index_permissions(tmp_path):
+    test_dir = tmp_path / "test"
+    with mock.patch.object(learn, "INDEX_SAVE_DIR", new=test_dir):
+        handler = MockLearnHandler()
+        handler._ensure_dirs()
+        mode = os.stat(test_dir).st_mode
+        assert stat.filemode(mode) == "drwx------"
+
+
 # TODO
 # import json