feat(agentchat): add get_thread() method to BaseGroupChat

[ayzmkk-rongbi]

Why are these changes needed?

Currently, the message thread is stored internally in BaseGroupChatManager._message_thread but is not accessible from the public API. Users need a way to:

• Inspect conversation progress during or after a group chat run
• Debug multi-agent interactions by examining message history
• Display chat history in UIs or monitoring dashboards
• Access the full message thread without relying on run_stream()

Related Issue(s)

• closes Get current message thread from a group chat team. #6085

Changes

_events.py: Added GroupChatGetThread RPC request event and GroupChatGetThreadResponse response model containing the serialized message list.

_base_group_chat_manager.py:

• Imported new event types
• Added handle_get_thread() RPC handler that returns a copy of self._message_thread

_base_group_chat.py:

• Imported new event types
• Added public async def get_thread() method with full documentation and usage example
• Handles both embedded and external runtime modes (starts/stops embedded runtime when team is not running)
• Validates team initialization before sending RPC

test_group_chat_get_thread.py: Added three test cases covering:

• test_get_thread_after_run: Verifies thread contains correct messages after run()
• test_get_thread_empty_after_reset: Verifies thread is empty after reset()
• test_get_thread_not_initialized: Verifies RuntimeError when team not initialized

All tests run with both single_threaded and embedded runtime modes (6 tests total, all passing).

Implementation Details

The implementation follows the existing RPC communication pattern used by reset(), pause(), and resume():

1. BaseGroupChat.get_thread() sends a GroupChatGetThread RPC message to the manager
2. BaseGroupChatManager.handle_get_thread() returns GroupChatGetThreadResponse with the current _message_thread
3. The response is deserialized and returned as List[BaseAgentEvent | BaseChatMessage]

For embedded runtime support, get_thread() temporarily starts the runtime if the team is not currently running (same pattern as reset()).

Usage Example

from autogen_agentchat.agents import AssistantAgent
from autogen_agentchat.conditions import MaxMessageTermination
from autogen_agentchat.teams import RoundRobinGroupChat
from autogen_ext.models.openai import OpenAIChatCompletionClient

async def main():
    model_client = OpenAIChatCompletionClient(model="gpt-4o")
    agent1 = AssistantAgent("Assistant1", model_client=model_client)
    agent2 = AssistantAgent("Assistant2", model_client=model_client)
    termination = MaxMessageTermination(3)
    team = RoundRobinGroupChat([agent1, agent2], termination_condition=termination)

    result = await team.run(task="Count from 1 to 10, respond one at a time.")

    # Get message history (new feature!)
    thread = await team.get_thread()
    for msg in thread:
        print(f"{msg.source}: {msg.content}")

[ayzmkk-rongbi]

@microsoft-github-policy-service agree company="Microsoft"

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 81.20%. Comparing base (13e144e) to head (c28fd95).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #7320      +/-   ##
==========================================
- Coverage   81.22%   81.20%   -0.02%     
==========================================
  Files         244        2     -242     
  Lines       18512      149   -18363     
==========================================
- Hits        15036      121   -14915     
+ Misses       3476       28    -3448     

Flag	Coverage Δ
unittests	?

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.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[ayzmkk-rongbi]

Hi @ekzhu, this PR implements the \get_thread()\ method for \BaseGroupChat\ as described in #6085. It follows the same RPC pattern used by
eset(), \pause(), and
esume(). All 6 new tests pass (3 test cases × 2 runtime modes). Would you be able to take a look when you have a chance? Thank you!