feat(integration): LangChain memory provider

[xe-nvdk]

Summary

Build a LangChain-compatible memory class (MemtraceMemory) that implements LangChain's BaseMemory or BaseChatMessageHistory interface. This lets any LangChain agent/chain use Memtrace as its memory backend.

Requirements

• Implement BaseMemory and/or BaseChatMessageHistory
• Support ConversationBufferMemory and ConversationSummaryMemory patterns
• Map session context endpoint to load_memory_variables()
• Publish as langchain-memtrace on PyPI

Dependencies

• Requires: Python SDK (feat(sdk): Python SDK #4)

Key Mapping

LangChain	Memtrace
load_memory_variables()	get_session_context()
save_context()	remember()
clear()	close_session()