feat: add Streamlit team example with stop-and-resume pattern

python/samples/agentchat_streamlit_team/README.md

@@ -0,0 +1,45 @@
+# Streamlit Team Chat Example (Stop-and-Resume Pattern)
+
+A Streamlit app demonstrating the **stop-and-resume** pattern for interactive multi-agent teams.
+
+Unlike the [chainlit example](../agentchat_chainlit/) which uses `UserProxyAgent` with a blocking input function, this example uses `MaxMessageTermination(max_messages=1)` to run the assistant for exactly one turn, then returns control to the Streamlit UI. This avoids the async blocking issue that Streamlit has with `UserProxyAgent.input_func`.
+
+## When to Use This Pattern
+
+- Web UIs built with Streamlit, Gradio, or similar frameworks
+- Any scenario where you can't block the main thread waiting for user input
+- When you need to save/resume team state across page reruns
+
+## Setup
+
+```bash
+pip install -r requirements.txt
+```
+
+Set your API key:
+```bash
+export OPENAI_API_KEY=sk-...
+```
+
+## Run
+
+```bash
+streamlit run app.py
+```
+
+## How It Works
+
+1. User sends a message via `st.chat_input`
+2. The team runs for **one turn** (`max_messages=1`) and stops
+3. The assistant's response is displayed
+4. The team state is preserved in `st.session_state` so the conversation continues
+5. When the user sends another message, the team resumes from where it left off
+
+## Alternative: UserProxyAgent with ChainLit
+
+If you need real-time human-in-the-loop interaction (where the agent asks questions mid-conversation), see the [ChainLit example](../agentchat_chainlit/) which supports `UserProxyAgent` natively.
+
+## Related
+
+- [Human-in-the-loop tutorial](https://microsoft.github.io/autogen/stable/user-guide/agentchat-user-guide/tutorial/human-in-the-loop.html)
+- [Termination conditions](https://microsoft.github.io/autogen/stable/user-guide/agentchat-user-guide/tutorial/termination.html)

python/samples/agentchat_streamlit_team/app.py

@@ -0,0 +1,80 @@
+# Copyright (c) Microsoft. All rights reserved.
+# Streamlit example: Team with AssistantAgent using stop-and-resume pattern.
+
+import asyncio
+import yaml
+import streamlit as st
+
+from autogen_agentchat.agents import AssistantAgent
+from autogen_agentchat.conditions import MaxMessageTermination, TextMentionTermination
+from autogen_agentchat.teams import RoundRobinGroupChat
+from autogen_core.models import ChatCompletionClient
+
+
+def init_session():
+    """Initialize Streamlit session state."""
+    if "messages" not in st.session_state:
+        st.session_state["messages"] = []
+    if "model_client" not in st.session_state:
+        with open("model_config.yaml", "r") as f:
+            config = yaml.safe_load(f)
+        st.session_state["model_client"] = ChatCompletionClient.load_component(config)
+    if "team" not in st.session_state:
+        assistant = AssistantAgent(
+            name="assistant",
+            model_client=st.session_state["model_client"],
+            system_message="You are a helpful AI assistant. Keep responses concise.",
+        )
+        termination = MaxMessageTermination(max_messages=1)
+        st.session_state["team"] = RoundRobinGroupChat(
+            [assistant],
+            termination_condition=termination,
+        )
+
+
+async def run_team(task: str):
+    """Run the team for one turn and return the result."""
+    result = await st.session_state["team"].run(task=task)
+    return result
+
+
+def main():
+    st.set_page_config(page_title="Team Chat (Stop-and-Resume)", page_icon="🤖")
+    st.title("Team Chat: Stop-and-Resume Pattern 🤖")
+
+    init_session()
+
+    # Display chat history
+    for msg in st.session_state["messages"]:
+        with st.chat_message(msg["role"]):
+            st.markdown(msg["content"])
+
+    prompt = st.chat_input("Type a message...")
+    if prompt:
+        # Show user message
+        st.session_state["messages"].append({"role": "user", "content": prompt})
+        with st.chat_message("user"):
+            st.markdown(prompt)
+
+        # Run team for one turn using a new event loop
+        loop = asyncio.new_event_loop()
+        try:
+            result = loop.run_until_complete(run_team(prompt))
+        finally:
+            loop.close()
+
+        # Display assistant response
+        for msg in result.messages:
+            if hasattr(msg, "content") and msg.source == "assistant":
+                content = msg.content
+                if content and content != prompt:
+                    st.session_state["messages"].append(
+                        {"role": "assistant", "content": content}
+                    )
+                    with st.chat_message("assistant"):
+                        st.markdown(content)
+                    break  # Only show the first assistant response
+
+
+if __name__ == "__main__":
+    main()

python/samples/agentchat_streamlit_team/model_config.yaml

@@ -0,0 +1,3 @@
+model: gpt-4o-mini
+api_type: openai
+# Set your API key via environment variable: export OPENAI_API_KEY=sk-...

python/samples/agentchat_streamlit_team/requirements.txt

@@ -0,0 +1,4 @@
+autogen-agentchat>=0.4
+autogen-ext[openai]>=0.4
+streamlit>=1.30
+pyyaml