What would you like to see?

Description:
First of all, thank you for the amazing work on AnythingLLM! The recent addition of RTL (Right-to-Left) support is a great improvement for Arabic users. However, there is still one issue that makes reading conversations difficult:

🛑 Problem:
While user messages are correctly aligned from right to left, the model responses are still aligned from left to right. This inconsistency creates a frustrating experience for Arabic speakers, especially when a response contains both Arabic and English words. This misalignment can even make sentences unreadable in some cases.

✅ Proposed Solution:

Ensure that model responses are also aligned from right to left when RTL mode is active.
This would greatly improve the user experience for Arabic users and make the conversation more natural and readable.
🎯 Why It Matters:
Arabic is spoken by hundreds of millions of users worldwide, and proper RTL support would make AnythingLLM much more accessible to them. Many Arabic users rely on LLMs for work, learning, and research, and improving readability would enhance their productivity.

We really appreciate your efforts, and we hope you consider this small but very impactful change. Thank you again for your hard work! 😊🚀