fix: revert screenshots to side-by-side table layout

Author: simba

README.md

@@ -37,21 +37,20 @@
 
 </div>
 
-<div align="center">
-
-**Browse & Run VLMs Locally from HuggingFace**
-
-<img src="screenshots/aegis-vlm-browser.png" alt="Aegis AI — Browse and download vision language models from HuggingFace" width="80%">
-
-*Download and run SmolVLM2, Qwen-VL, LFM2.5, LLaVA, MiniCPM-V directly on your machine. Even a Mac M1 Mini 8GB works.*
-
-**Chat with your AI Security Agent**
-
-<img src="screenshots/aegis-chat-agent.png" alt="Aegis AI — Ask your agent what happened and get real answers" width="80%">
-
-*Ask "anyone entering the room?" — Aegis searches your footage and gives you a real answer with timestamps and clips.*
-
-</div>
+<table>
+<tr>
+<td width="50%">
+<p align="center"><b>Browse & Run VLMs Locally from HuggingFace</b></p>
+<img src="screenshots/aegis-vlm-browser.png" alt="Aegis AI — Browse and download vision language models from HuggingFace" width="100%">
+<p align="center"><em>Download and run SmolVLM2, Qwen-VL, LFM2.5, LLaVA, MiniCPM-V directly on your machine. Even a Mac M1 Mini 8GB works.</em></p>
+</td>
+<td width="50%">
+<p align="center"><b>Chat with your AI Security Agent</b></p>
+<img src="screenshots/aegis-chat-agent.png" alt="Aegis AI — Ask your agent what happened and get real answers" width="100%">
+<p align="center"><em>Ask "anyone entering the room?" — Aegis searches your footage and gives you a real answer with timestamps and clips.</em></p>
+</td>
+</tr>
+</table>
 
 ---