feat: support for async llm callback

[OpenGaet]

Issue #, if available:

Description of changes:

Added Async LLM Callback Support

New Features:
• Added process_turn_with_llm_async() method to MemorySessionManager and MemorySession classes
• Supports async LLM callbacks that return Awaitable[str]
• Maintains full backwards compatibility with existing sync callbacks

Code Improvements:
• Extracted common logic into helper methods (_retrieve_memories_for_llm() and _save_conversation_turn())
• Added test coverage for async functionality

Files Modified:
• src/bedrock_agentcore/memory/session.py (+122/-16 lines)
• tests/bedrock_agentcore/memory/test_session.py (+35 lines)

Usage:

# Async callback support
await manager.process_turn_with_llm_async(                                                                                                                                                                         
    actor_id="user-123",                                                                                                                                                                                           
    session_id="session-456",                                                                                                                                                                                      
    user_input="Hello",                                                                                                                                                                                            
    llm_callback=async_callback,  # async function                                                                                                                                                                 
    retrieval_config=config                                                                                                                                                                                        
)

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[bergjaak]

What if instead, we make it so _retrieve_memories_for_llm requires the retrieval_config but then this if retrieval_config check happens where we call _retrieve_memories_for_llm? I think that would be a less accident prone API since then there is no way the user tries to call _retrieve_memories_for_llm without a retrieval config and is surprised that they get no memory records

[OpenGaet]

This was the existing logic, just moved to a separate function to reduce code duplication. On top of this, it is a "private" function, so the user should not be calling _retrieve_memories_for_llm directly. But feel free to make changes as desired.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 80.64516% with 6 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (main@f30e281). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
src/bedrock_agentcore/memory/session.py	80.64%	5 Missing and 1 partial ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #131   +/-   ##
=======================================
  Coverage        ?   89.39%           
=======================================
  Files           ?       24           
  Lines           ?     2414           
  Branches        ?      342           
=======================================
  Hits            ?     2158           
  Misses          ?      151           
  Partials        ?      105           

Flag	Coverage Δ
unittests	89.39% <80.64%> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.