docs: update LangGraph cookbook for current LangChain/LangGraph APIs (#3103)

Co-authored-by: Hassieb Pakzad <68423100+hassiebp@users.noreply.github.com>

Author: milanagm

content/guides/cookbook/integration_langgraph.mdx

@@ -228,17 +228,24 @@ from typing import Annotated
 from langchain_community.tools import WikipediaQueryRun
 from langchain_community.utilities import WikipediaAPIWrapper
 from datetime import datetime
-from langchain.tools import Tool
+from langchain.tools import tool
 
-# Define a tools that searches Wikipedia
-wikipedia_tool = WikipediaQueryRun(api_wrapper=WikipediaAPIWrapper())
+# Define a tool that searches Wikipedia
+_wikipedia_query = WikipediaQueryRun(api_wrapper=WikipediaAPIWrapper())
+
+@tool("wikipedia")
+def wikipedia_tool(query: str) -> str:
+    """Search Wikipedia for the given query and return a summary."""
+    try:
+        return _wikipedia_query.invoke(query)
+    except Exception as exc:
+        return f"Wikipedia lookup failed ({exc}). Answer from your own knowledge instead."
 
 # Define a new tool that returns the current datetime
-datetime_tool = Tool(
-    name="Datetime",
-    func = lambda x: datetime.now().isoformat(),
-    description="Returns the current datetime",
-)
+@tool("current_datetime")
+def datetime_tool() -> str:
+    """Returns the current datetime."""
+    return datetime.now().isoformat()
 ```
 
 ### Helper utilities
@@ -247,29 +254,21 @@ Define a helper function below to simplify adding new agent worker nodes.
 
 
 ```python
-from langchain.agents import AgentExecutor, create_openai_tools_agent
+from langchain.agents import create_agent
 from langchain_core.messages import BaseMessage, HumanMessage
 from langchain_openai import ChatOpenAI
 
-def create_agent(llm: ChatOpenAI, system_prompt: str, tools: list):
+def create_worker_agent(llm: ChatOpenAI, system_prompt: str, tools: list):
     # Each worker node will be given a name and some tools.
-    prompt = ChatPromptTemplate.from_messages(
-        [
-            (
-                "system",
-                system_prompt,
-            ),
-            MessagesPlaceholder(variable_name="messages"),
-            MessagesPlaceholder(variable_name="agent_scratchpad"),
-        ]
+    return create_agent(
+        model=llm,
+        tools=tools,
+        system_prompt=system_prompt,
     )
-    agent = create_openai_tools_agent(llm, tools, prompt)
-    executor = AgentExecutor(agent=agent, tools=tools)
-    return executor
 
 def agent_node(state, agent, name):
     result = agent.invoke(state)
-    return {"messages": [HumanMessage(content=result["output"], name=name)]}
+    return {"messages": [HumanMessage(content=result["messages"][-1].content, name=name)]}
 ```
 
 ### Create agent supervisor
@@ -278,7 +277,6 @@ It will use function calling to choose the next worker node OR finish processing
 
 
 ```python
-from langchain_core.output_parsers.openai_functions import JsonOutputFunctionsParser
 from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
 
 members = ["Researcher", "CurrentTime"]
@@ -292,7 +290,7 @@ system_prompt = (
 # Our team supervisor is an LLM node. It just picks the next agent to process and decides when the work is completed
 options = ["FINISH"] + members
 
-# Using openai function calling can make output parsing easier for us
+# Using structured output can make routing easier for us
 function_def = {
     "name": "route",
     "description": "Select the next role.",
@@ -327,11 +325,7 @@ prompt = ChatPromptTemplate.from_messages(
 llm = ChatOpenAI(model="gpt-4o")
 
 # Construction of the chain for the supervisor agent
-supervisor_chain = (
-    prompt
-    | llm.bind_functions(functions=[function_def], function_call="route")
-    | JsonOutputFunctionsParser()
-)
+supervisor_chain = prompt | llm.with_structured_output(function_def)
 ```
 
 ### Construct graph
@@ -354,11 +348,11 @@ class AgentState(TypedDict):
     next: str
 
 # Add the research agent using the create_agent helper function
-research_agent = create_agent(llm, "You are a web researcher.", [wikipedia_tool])
+research_agent = create_worker_agent(llm, "You are a web researcher.", [wikipedia_tool])
 research_node = functools.partial(agent_node, agent=research_agent, name="Researcher")
 
 # Add the time agent using the create_agent helper function
-currenttime_agent = create_agent(llm, "You can tell the current time at", [datetime_tool])
+currenttime_agent = create_worker_agent(llm, "You can tell the current time.", [datetime_tool])
 currenttime_node = functools.partial(agent_node, agent=currenttime_agent, name = "CurrentTime")
 
 workflow = StateGraph(AgentState)

cookbook/integration_langgraph.ipynb

@@ -396,17 +396,24 @@
     "from langchain_community.tools import WikipediaQueryRun\n",
     "from langchain_community.utilities import WikipediaAPIWrapper\n",
     "from datetime import datetime\n",
-    "from langchain.tools import Tool\n",
+    "from langchain.tools import tool\n",
     "\n",
-    "# Define a tools that searches Wikipedia\n",
-    "wikipedia_tool = WikipediaQueryRun(api_wrapper=WikipediaAPIWrapper())\n",
+    "# Define a tool that searches Wikipedia\n",
+    "_wikipedia_query = WikipediaQueryRun(api_wrapper=WikipediaAPIWrapper())\n",
+    "\n",
+    "@tool(\"wikipedia\")\n",
+    "def wikipedia_tool(query: str) -> str:\n",
+    "    \"\"\"Search Wikipedia for the given query and return a summary.\"\"\"\n",
+    "    try:\n",
+    "        return _wikipedia_query.invoke(query)\n",
+    "    except Exception as exc:\n",
+    "        return f\"Wikipedia lookup failed ({exc}). Answer from your own knowledge instead.\"\n",
     "\n",
     "# Define a new tool that returns the current datetime\n",
-    "datetime_tool = Tool(\n",
-    "    name=\"Datetime\",\n",
-    "    func = lambda x: datetime.now().isoformat(),\n",
-    "    description=\"Returns the current datetime\",\n",
-    ")"
+    "@tool(\"current_datetime\")\n",
+    "def datetime_tool() -> str:\n",
+    "    \"\"\"Returns the current datetime.\"\"\"\n",
+    "    return datetime.now().isoformat()"
    ]
   },
   {
@@ -428,29 +435,21 @@
    },
    "outputs": [],
    "source": [
-    "from langchain.agents import AgentExecutor, create_openai_tools_agent\n",
+    "from langchain.agents import create_agent\n",
     "from langchain_core.messages import BaseMessage, HumanMessage\n",
     "from langchain_openai import ChatOpenAI\n",
     "\n",
-    "def create_agent(llm: ChatOpenAI, system_prompt: str, tools: list):\n",
+    "def create_worker_agent(llm: ChatOpenAI, system_prompt: str, tools: list):\n",
     "    # Each worker node will be given a name and some tools.\n",
-    "    prompt = ChatPromptTemplate.from_messages(\n",
-    "        [\n",
-    "            (\n",
-    "                \"system\",\n",
-    "                system_prompt,\n",
-    "            ),\n",
-    "            MessagesPlaceholder(variable_name=\"messages\"),\n",
-    "            MessagesPlaceholder(variable_name=\"agent_scratchpad\"),\n",
-    "        ]\n",
+    "    return create_agent(\n",
+    "        model=llm,\n",
+    "        tools=tools,\n",
+    "        system_prompt=system_prompt,\n",
     "    )\n",
-    "    agent = create_openai_tools_agent(llm, tools, prompt)\n",
-    "    executor = AgentExecutor(agent=agent, tools=tools)\n",
-    "    return executor\n",
     "\n",
     "def agent_node(state, agent, name):\n",
     "    result = agent.invoke(state)\n",
-    "    return {\"messages\": [HumanMessage(content=result[\"output\"], name=name)]}"
+    "    return {\"messages\": [HumanMessage(content=result[\"messages\"][-1].content, name=name)]}"
    ]
   },
   {
@@ -472,7 +471,6 @@
    },
    "outputs": [],
    "source": [
-    "from langchain_core.output_parsers.openai_functions import JsonOutputFunctionsParser\n",
     "from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder\n",
     "\n",
     "members = [\"Researcher\", \"CurrentTime\"]\n",
@@ -486,7 +484,7 @@
     "# Our team supervisor is an LLM node. It just picks the next agent to process and decides when the work is completed\n",
     "options = [\"FINISH\"] + members\n",
     "\n",
-    "# Using openai function calling can make output parsing easier for us\n",
+    "# Using structured output can make routing easier for us\n",
     "function_def = {\n",
     "    \"name\": \"route\",\n",
     "    \"description\": \"Select the next role.\",\n",
@@ -521,11 +519,7 @@
     "llm = ChatOpenAI(model=\"gpt-4o\")\n",
     "\n",
     "# Construction of the chain for the supervisor agent\n",
-    "supervisor_chain = (\n",
-    "    prompt\n",
-    "    | llm.bind_functions(functions=[function_def], function_call=\"route\")\n",
-    "    | JsonOutputFunctionsParser()\n",
-    ")"
+    "supervisor_chain = prompt | llm.with_structured_output(function_def)"
    ]
   },
   {
@@ -561,11 +555,11 @@
     "    next: str\n",
     "\n",
     "# Add the research agent using the create_agent helper function\n",
-    "research_agent = create_agent(llm, \"You are a web researcher.\", [wikipedia_tool])\n",
+    "research_agent = create_worker_agent(llm, \"You are a web researcher.\", [wikipedia_tool])\n",
     "research_node = functools.partial(agent_node, agent=research_agent, name=\"Researcher\")\n",
     "\n",
     "# Add the time agent using the create_agent helper function\n",
-    "currenttime_agent = create_agent(llm, \"You can tell the current time at\", [datetime_tool])\n",
+    "currenttime_agent = create_worker_agent(llm, \"You can tell the current time.\", [datetime_tool])\n",
     "currenttime_node = functools.partial(agent_node, agent=currenttime_agent, name = \"CurrentTime\")\n",
     "\n",
     "workflow = StateGraph(AgentState)\n",