merge: upstream/experimental with citations feature

- Merge upstream changes including image search, tooltips, and UI improvements
- Keep citations feature with inline hover cards
- Resolve conflict in message-list-item.tsx: use upstream img max-width (90%) while preserving citations logic
- Maintain file upload improvements with citations support

Co-authored-by: Cursor <cursoragent@cursor.com>

backend/src/agents/lead_agent/prompt.py

@@ -158,6 +158,7 @@ The key AI trends for 2026 include enhanced reasoning capabilities, multimodal i
 - Progressive Loading: Load resources incrementally as referenced in skills
 - Output Files: Final deliverables must be in `/mnt/user-data/outputs`
 - Clarity: Be direct and helpful, avoid unnecessary meta-commentary
+- Including Images and Mermaid: Images and Mermaid diagrams are always welcomed in the Markdown format, and you're encouraged to use `![Image Description](image_path)\n\n` or "```mermaid" to display images in response or Markdown files
 - Multi-task: Better utilize parallel tool calling to call multiple tools at one time for better performance
 - Language Consistency: Keep using the same language as user's
 - Always Respond: Your thinking is internal. You MUST always provide a visible response to the user after thinking.

backend/src/agents/middlewares/clarification_middleware.py

@@ -5,7 +5,7 @@ from typing import override
 
 from langchain.agents import AgentState
 from langchain.agents.middleware import AgentMiddleware
-from langchain_core.messages import AIMessage, ToolMessage
+from langchain_core.messages import ToolMessage
 from langgraph.graph import END
 from langgraph.prebuilt.tool_node import ToolCallRequest
 from langgraph.types import Command
@@ -118,17 +118,13 @@ class ClarificationMiddleware(AgentMiddleware[ClarificationMiddlewareState]):
             name="ask_clarification",
         )
 
-        ai_response_message = AIMessage(content=formatted_message)
-
         # Return a Command that:
-        # 1. Adds the formatted tool message (keeping the AI message intact)
+        # 1. Adds the formatted tool message
         # 2. Interrupts execution by going to __end__
-        # Note: We don't modify the AI message to preserve all fields (reasoning_content, tool_calls, etc.)
-        # This is especially important for thinking mode where reasoning_content is required
-
-        # Return Command to add the tool message and interrupt
+        # Note: We don't add an extra AIMessage here - the frontend will detect
+        # and display ask_clarification tool messages directly
         return Command(
-            update={"messages": [tool_message, ai_response_message]},
+            update={"messages": [tool_message]},
             goto=END,
         )