add simple multi agent for explaining Command function

docs/references.md

@@ -9,6 +9,9 @@
 - [Streamlit](https://python.langchain.com/docs/integrations/callbacks/streamlit/)
 - [LangChain MCP Adapters](https://github.com/langchain-ai/langchain-mcp-adapters)
   - [Research Agent with MCP Integration.](https://github.com/langchain-ai/deep_research_from_scratch/blob/main/src/deep_research_from_scratch/research_agent_mcp.py)
+- [Command: A new tool for building multi-agent architectures in LangGraph](https://blog.langchain.com/command-a-new-tool-for-multi-agent-architectures-in-langgraph/)
+- [Combine control flow and state updates with Command](https://langchain-ai.github.io/langgraph/how-tos/graph-api/#combine-control-flow-and-state-updates-with-command)
+- [Command: a new tool for building multi-agent architectures in LangGraph](https://www.youtube.com/watch?v=6BJDKf90L9A)
 
 ### Sample Codes
 

pyproject.toml

@@ -85,3 +85,4 @@ non-subscriptable = "ignore"
 possibly-unbound-attribute = "ignore"
 unresolved-attribute = "ignore"
 invalid-argument-type = "ignore"
+invalid-type-form = "ignore"

scripts/simple_multi_agent_operator.py

@@ -0,0 +1,80 @@
+import logging
+
+import typer
+from dotenv import load_dotenv
+
+from template_langgraph.agents.simple_multi_agent.multi_agent import app as multi_agent_app
+from template_langgraph.agents.simple_multi_agent.weather_agent import app as weather_agent_app
+from template_langgraph.loggers import get_logger
+
+app = typer.Typer(
+    add_completion=False,
+    help="SimpleMultiAgent CLI",
+)
+logger = get_logger(__name__)
+
+
+@app.command()
+def weather_agent(
+    query: str = typer.Option(
+        "What's the weather in Japan?",
+        "--query",
+        "-q",
+        help="The query to ask the model",
+    ),
+    verbose: bool = typer.Option(
+        False,
+        "--verbose",
+        "-v",
+        help="Enable verbose output",
+    ),
+):
+    if verbose:
+        logger.setLevel(logging.DEBUG)
+
+    response = weather_agent_app.invoke(
+        {
+            "messages": [
+                {"role": "user", "content": query},
+            ],
+        },
+        debug=True,
+    )
+    logger.info(response["messages"][-1].content)
+
+
+@app.command()
+def multi_agent(
+    query: str = typer.Option(
+        "What's the weather in Japan?",
+        "--query",
+        "-q",
+        help="The query to ask the model",
+    ),
+    verbose: bool = typer.Option(
+        False,
+        "--verbose",
+        "-v",
+        help="Enable verbose output",
+    ),
+):
+    if verbose:
+        logger.setLevel(logging.DEBUG)
+
+    response = multi_agent_app.invoke(
+        {
+            "messages": [
+                {"role": "user", "content": query},
+            ],
+        },
+        debug=True,
+    )
+    logger.info(response["messages"][-1].content)
+
+
+if __name__ == "__main__":
+    load_dotenv(
+        override=True,
+        verbose=True,
+    )
+    app()

template_langgraph/agents/simple_multi_agent/multi_agent.py

@@ -0,0 +1,42 @@
+from typing import Literal
+
+from langgraph.graph import END, START, MessagesState, StateGraph
+from langgraph.types import Command
+
+from template_langgraph.agents.simple_multi_agent.weather_agent import app
+from template_langgraph.llms.azure_openais import AzureOpenAiWrapper
+from template_langgraph.loggers import get_logger
+
+logger = get_logger(__name__)
+
+
+def transfer_to_weather_agent():
+    """Call this to transfer to the weather agent"""
+
+
+tools = [transfer_to_weather_agent]
+llm = AzureOpenAiWrapper().chat_model.bind_tools(tools=tools)
+
+
+def call_model(state: MessagesState) -> Command[Literal["weather_agent", END]]:
+    messages = state["messages"]
+    response = llm.invoke(messages)
+    if len(response.tool_calls) > 0:
+        return Command(
+            goto="weather_agent",
+        )
+    else:
+        return Command(
+            goto=END,
+            update={
+                "messages": [response],
+            },
+        )
+
+
+workflow = StateGraph(MessagesState)
+
+workflow.add_node("agent", call_model)
+workflow.add_node("weather_agent", app)
+workflow.add_edge(START, "agent")
+app = workflow.compile()

template_langgraph/agents/simple_multi_agent/weather_agent.py

@@ -0,0 +1,45 @@
+from typing import Literal
+
+from langchain_core.tools import tool
+from langgraph.graph import END, START, MessagesState, StateGraph
+from langgraph.prebuilt import ToolNode
+from langgraph.types import Command
+
+from template_langgraph.llms.azure_openais import AzureOpenAiWrapper
+
+
+@tool
+def search(query: str) -> str:
+    """Call to surf the web"""
+    if "japan" in query.lower():
+        return "It's 60 degrees and cloudy in Japan"
+    return "It's 90 degrees and sunny in Japan"
+
+
+tools = [search]
+tool_node = ToolNode(tools=tools)
+llm = AzureOpenAiWrapper().chat_model.bind_tools(tools=tools)
+
+
+def call_model(state: MessagesState) -> Command[Literal["tools", END]]:
+    messages = state["messages"]
+    response = llm.invoke(messages)
+    if len(response.tool_calls) > 0:
+        next_node = "tools"
+    else:
+        next_node = END
+    return Command(
+        goto=next_node,
+        update={
+            "messages": [response],
+        },
+    )
+
+
+workflow = StateGraph(MessagesState)
+
+workflow.add_node("agent", call_model)
+workflow.add_node("tools", tool_node)
+workflow.add_edge(START, "agent")
+workflow.add_edge("tools", "agent")
+app = workflow.compile()