refactor: remove welcome message and prevent chat repopulation on model switch

Co-authored-by: aider (openai//models/qwen2.5-coder-32b-instruct-q4_k_m.gguf) <aider@aider.chat>

chatv2.js

@@ -613,33 +613,6 @@ window.onload = () => {
 
 	// Load chat history from localStorage
 	if (chatHistory.length === 0) {
-		// Add welcome message (this will be the first entry in chatHistory)
-		const welcomeMessage = `# Welcome to LM Studio Chat!
-
-I now support **full conversation history**, **real-time streaming responses** and **dark, readable text formatting**.
-
-## Features:
-1. **Full Chat History** - Complete conversation context is now sent to the AI
-2. **Smart History Management** - Automatically keeps the last ${ MAX_CHAT_HISTORY } messages
-3. **Streaming Mode** (enabled by default) - Watch responses appear word-by-word
-4. **Markdown Rendering** - Proper formatting for code, lists, tables, and more
-5. **Readable Dark Text** - No more eye strain from light gray text
-
-## Try it out:
-- Ask follow-up questions that reference earlier messages
-- Have a multi-turn conversation with full context
-- Ask "what did I just ask?" to test history retention
-- Watch the streaming response in real-time!
-
-> *Tip: You can toggle streaming and markdown using the checkboxes below.*`
-
-		addMessage('assistant', welcomeMessage, true)
-	} else {
-		// Re-add all messages from chatHistory
-		chatHistory.forEach(message => {
-			addMessage(message.role, message.content, message.markdown, message.messageId)
-		})
-	}
 
 	// Add event listener for reset button
 	document.getElementById('resetBtn').addEventListener('click', resetChat)