Add multi-part content support to preload_memory_tool

[sezerogras]

This PR adds support for multi-part content in the PreloadMemoryTool. Previously, the tool only supported text content, but now it can handle:

• Text parts
• Function calls
• Function responses
• Inline data (like images)

Each content type is formatted in a human-readable way in the memory context.

Resolves the TODO: "support multi-part content" in preload_memory_tool.py

[hangfei]

I believe this is already supported. Please update to the latest to retry. Feel free to open a new issue/PR if you still see the gap.