Lord of Large Language Models (LoLLMs) path traversal vulnerability in the api open_personality_folder endpoint

A path traversal vulnerability exists in the api open_personality_folder endpoint of parisneo/lollms. This vulnerability allows an attacker to read any folder in the personality_folder on the victim's computer, even though sanitize_path is set. The issue arises due to improper sanitization of the personality_folder parameter, which can be exploited to traverse directories and access arbitrary files.