Is your feature request related to a problem? Please describe.
Two related pagination issues affect developer experience:

1. get_last_k_turns(k=200) - Requesting 200 turns only fetches 100 events (default max_results), so users don't get the expected number of turns. k and max_results are independent parameters, which is unintuitive.

# User expects 200 turns, but only gets turns found within 100 events
turns = client.get_last_k_turns(k=200)  # max_results=100 by default

2. list_messages(limit=None) - Returns 100 messages instead of all messages. Users must guess a "large enough" number to fetch complete history.

# Expected: all messages, Actual: 100 messages
messages = session_repository.list_messages()

# Current workaround - arbitrary large number
messages = session_repository.list_messages(limit=10000)

Describe the solution you'd like

1. get_last_k_turns(k) in memory/session.py, memory/client.py: Auto-calculate max_results based on k if not explicitly specified

def get_last_k_turns(self, ..., k: int = 5, max_results: Optional[int] = None):
    if max_results is None:
        max_results = max(100, k * 3)  # estimate ~2-3 events per turn
    events = self.list_events(max_results=max_results)
      ...

2. list_messages() in memory/integrations/strands/session_manager.py: Add fetch_all parameter for explicit intent.

messages = list_messages()                # 100 (default, backward compatible)
messages = list_messages(limit=500)       # 500
messages = list_messages(fetch_all=True)  # all messages

Describe alternatives you've considered

1. Change limit=None to mean "fetch all" - Follows common API conventions, but may break existing code that relies on current default behavior (max 100).
2. Increase default max_results to a larger value (e.g., 1000) - Simple but doesn't solve the core issue and may cause performance problems.
3. Document current behavior more clearly - Doesn't improve DevX, just shifts burden to users.

Additional context
Use cases affected:

• Session restoration requiring complete conversation history
• Context compaction / summarization
• Conversation export
• Analytics