langgraph/examples/chat_agent_executor_with_function_calling/human-in-the-loop.ipynb

Human-in-the-loop

In this example we will build a chat executor that has a human in the loop. We will use the human to approve specific actions.

This examples builds off the base chat executor. It is highly recommended you learn about that executor before going through this notebook. You can find documentation for that example here.

Any modifications of that example are called below with MODIFICATION, so if you are looking for the differences you can just search for that.

Setup

First we need to install the packages required

In [1]:

!pip install --quiet -U langchain langchain_openai tavily-python

Next, we need to set API keys for OpenAI (the LLM we will use) and Tavily (the search tool we will use)

In [ ]:

import os
import getpass

os.environ["OPENAI_API_KEY"] = getpass.getpass("OpenAI API Key:")
os.environ["TAVILY_API_KEY"] = getpass.getpass("Tavily API Key:")

Optionally, we can set API key for LangSmith tracing, which will give us best-in-class observability.

In [ ]:

os.environ["LANGCHAIN_TRACING_V2"] = "true"
os.environ["LANGCHAIN_API_KEY"] = getpass.getpass("LangSmith API Key:")

Set up the tools

We will first define the tools we want to use. For this simple example, we will use a built-in search tool via Tavily. However, it is really easy to create your own tools - see documentation here on how to do that.

In [1]:

from langchain_community.tools.tavily_search import TavilySearchResults

tools = [TavilySearchResults(max_results=1)]

We can now wrap these tools in a simple ToolExecutor. This is a real simple class that takes in a ToolInvocation and calls that tool, returning the output. A ToolInvocation is any class with tool and tool_input attribute.

In [2]:

from langgraph.prebuilt import ToolExecutor

tool_executor = ToolExecutor(tools)

Set up the model

Now we need to load the chat model we want to use. Importantly, this should satisfy two criteria:

1. It should work with messages. We will represent all agent state in the form of messages, so it needs to be able to work well with them.
2. It should work with OpenAI function calling. This means it should either be an OpenAI model or a model that exposes a similar interface.

Note: these model requirements are not requirements for using LangGraph - they are just requirements for this one example.

In [3]:

from langchain_openai import ChatOpenAI

# We will set streaming=True so that we can stream tokens
# See the streaming section for more information on this.
model = ChatOpenAI(temperature=0, streaming=True)

After we've done this, we should make sure the model knows that it has these tools available to call. We can do this by converting the LangChain tools into the format for OpenAI function calling, and then bind them to the model class.

In [4]:

from langchain.tools.render import format_tool_to_openai_function

functions = [format_tool_to_openai_function(t) for t in tools]
model = model.bind_functions(functions)

Define the agent state

The main type of graph in langgraph is the StatefulGraph. This graph is parameterized by a state object that it passes around to each node. Each node then returns operations to update that state. These operations can either SET specific attributes on the state (e.g. overwrite the existing values) or ADD to the existing attribute. Whether to set or add is denoted by annotating the state object you construct the graph with.

For this example, the state we will track will just be a list of messages. We want each node to just add messages to that list. Therefore, we will use a TypedDict with one key (messages) and annotate it so that the messages attribute is always added to.

In [5]:

from typing import TypedDict, Annotated, Sequence
import operator
from langchain_core.messages import BaseMessage


class AgentState(TypedDict):
    messages: Annotated[Sequence[BaseMessage], operator.add]

Define the nodes

We now need to define a few different nodes in our graph. In langgraph, a node can be either a function or a runnable. There are two main nodes we need for this:

1. The agent: responsible for deciding what (if any) actions to take.
2. A function to invoke tools: if the agent decides to take an action, this node will then execute that action.

We will also need to define some edges. Some of these edges may be conditional. The reason they are conditional is that based on the output of a node, one of several paths may be taken. The path that is taken is not known until that node is run (the LLM decides).

1. Conditional Edge: after the agent is called, we should either: a. If the agent said to take an action, then the function to invoke tools should be called b. If the agent said that it was finished, then it should finish
2. Normal Edge: after the tools are invoked, it should always go back to the agent to decide what to do next

Let's define the nodes, as well as a function to decide how what conditional edge to take.

In [6]:

from langgraph.prebuilt import ToolInvocation
import json
from langchain_core.messages import FunctionMessage

# Define the function that determines whether to continue or not
def should_continue(state):
    messages = state['messages']
    last_message = messages[-1]
    # If there is no function call, then we finish
    if "function_call" not in last_message.additional_kwargs:
        return "end"
    # Otherwise if there is, we continue
    else:
        return "continue"

# Define the function that calls the model
def call_model(state):
    messages = state['messages']
    response = model.invoke(messages)
    # We return a list, because this will get added to the existing list
    return {"messages": [response]}

MODIFICATION

We modify the function that is calling the tool to first ask for user approval to continue. Note that this is a simple example and we could modify it to change the tool input, use some other channel besides input, etc.

In [7]:

# Define the function to execute tools
def call_tool(state):
    messages = state['messages']
    # Based on the continue condition
    # we know the last message involves a function call
    last_message = messages[-1]
    # We construct an ToolInvocation from the function_call
    action = ToolInvocation(
        tool=last_message.additional_kwargs["function_call"]["name"],
        tool_input=json.loads(last_message.additional_kwargs["function_call"]["arguments"]),
    )
    response = input(prompt=f"[y/n] continue with: {action}?")
    if response == "n":
        raise ValueError
    # We call the tool_executor and get back a response
    response = tool_executor.invoke(action)
    # We use the response to create a FunctionMessage
    function_message = FunctionMessage(content=str(response), name=action.tool)
    # We return a list, because this will get added to the existing list
    return {"messages": [function_message]}

Define the graph

We can now put it all together and define the graph!

In [8]:

from langgraph.graph import StateGraph, END
# Define a new graph
workflow = StateGraph(AgentState)

# Define the two nodes we will cycle between
workflow.add_node("agent", call_model)
workflow.add_node("action", call_tool)

# Set the entrypoint as `agent`
# This means that this node is the first one called
workflow.set_entry_point("agent")

# We now add a conditional edge
workflow.add_conditional_edges(
    # First, we define the start node. We use `agent`.
    # This means these are the edges taken after the `agent` node is called.
    "agent",
    # Next, we pass in the function that will determine which node is called next.
    should_continue,
    # Finally we pass in a mapping.
    # The keys are strings, and the values are other nodes.
    # END is a special node marking that the graph should finish.
    # What will happen is we will call `should_continue`, and then the output of that
    # will be matched against the keys in this mapping.
    # Based on which one it matches, that node will then be called.
    {
        # If `tools`, then we call the tool node.
        "continue": "action",
        # Otherwise we finish.
        "end": END
    }
)

# We now add a normal edge from `tools` to `agent`.
# This means that after `tools` is called, `agent` node is called next.
workflow.add_edge('action', 'agent')

# Finally, we compile it!
# This compiles it into a LangChain Runnable,
# meaning you can use it as you would any other runnable
app = workflow.compile()

Use it!

We can now use it! This now exposes the same interface as all other LangChain runnables.

In [10]:

from langchain_core.messages import HumanMessage

inputs = {"messages": [HumanMessage(content="what is the weather in sf")]}
for output in app.stream(inputs):
    # stream() yields dictionaries with output keyed by node name
    for key, value in output.items():
        print(f"Output from node '{key}':")
        print("---")
        print(value)
    print("\n---\n")

Output from node 'agent':
---
{'messages': [AIMessage(content='', additional_kwargs={'function_call': {'arguments': '{\n  "query": "weather in San Francisco"\n}', 'name': 'tavily_search_results_json'}})]}

---

[y/n] continue with: tool='tavily_search_results_json' tool_input={'query': 'weather in San Francisco'}? n

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
Cell In[10], line 4
      1 from langchain_core.messages import HumanMessage
      3 inputs = {"messages": [HumanMessage(content="what is the weather in sf")]}
----> 4 for output in app.stream(inputs):
      5     # stream() yields dictionaries with output keyed by node name
      6     for key, value in output.items():
      7         print(f"Output from node '{key}':")

File ~/workplace/permchain/langgraph/pregel/__init__.py:528, in Pregel.transform(self, input, config, output_keys, input_keys, **kwargs)
    519 def transform(
    520     self,
    521     input: Iterator[Union[dict[str, Any], Any]],
   (...)
    526     **kwargs: Any,
    527 ) -> Iterator[Union[dict[str, Any], Any]]:
--> 528     for chunk in self._transform_stream_with_config(
    529         input,
    530         self._transform,
    531         config,
    532         output_keys=output_keys,
    533         input_keys=input_keys,
    534         **kwargs,
    535     ):
    536         yield chunk

File ~/workplace/langchain/libs/core/langchain_core/runnables/base.py:1226, in Runnable._transform_stream_with_config(self, input, transformer, config, run_type, **kwargs)
   1224 try:
   1225     while True:
-> 1226         chunk: Output = context.run(next, iterator)  # type: ignore
   1227         yield chunk
   1228         if final_output_supported:

File ~/workplace/permchain/langgraph/pregel/__init__.py:313, in Pregel._transform(self, input, run_manager, config, input_keys, output_keys)
    303 done, inflight = concurrent.futures.wait(
    304     [
    305         executor.submit(proc.invoke, input, config)
   (...)
    309     timeout=self.step_timeout,
    310 )
    312 # interrupt on failure or timeout
--> 313 _interrupt_or_proceed(done, inflight, step)
    315 # apply writes to channels
    316 _apply_writes(checkpoint, channels, pending_writes, config, step + 1)

File ~/workplace/permchain/langgraph/pregel/__init__.py:611, in _interrupt_or_proceed(done, inflight, step)
    609             inflight.pop().cancel()
    610         # raise the exception
--> 611         raise exc
    612         # TODO this is where retry of an entire step would happen
    614 if inflight:
    615     # if we got here means we timed out

File ~/.pyenv/versions/3.11.1/lib/python3.11/concurrent/futures/thread.py:58, in _WorkItem.run(self)
     55     return
     57 try:
---> 58     result = self.fn(*self.args, **self.kwargs)
     59 except BaseException as exc:
     60     self.future.set_exception(exc)

File ~/workplace/langchain/libs/core/langchain_core/runnables/base.py:3596, in RunnableBindingBase.invoke(self, input, config, **kwargs)
   3590 def invoke(
   3591     self,
   3592     input: Input,
   3593     config: Optional[RunnableConfig] = None,
   3594     **kwargs: Optional[Any],
   3595 ) -> Output:
-> 3596     return self.bound.invoke(
   3597         input,
   3598         self._merge_configs(config),
   3599         **{**self.kwargs, **kwargs},
   3600     )

File ~/workplace/langchain/libs/core/langchain_core/runnables/base.py:1774, in RunnableSequence.invoke(self, input, config)
   1772 try:
   1773     for i, step in enumerate(self.steps):
-> 1774         input = step.invoke(
   1775             input,
   1776             # mark each step as a child run
   1777             patch_config(
   1778                 config, callbacks=run_manager.get_child(f"seq:step:{i+1}")
   1779             ),
   1780         )
   1781 # finish the root run
   1782 except BaseException as e:

File ~/workplace/langchain/libs/core/langchain_core/runnables/base.py:3074, in RunnableLambda.invoke(self, input, config, **kwargs)
   3072 """Invoke this runnable synchronously."""
   3073 if hasattr(self, "func"):
-> 3074     return self._call_with_config(
   3075         self._invoke,
   3076         input,
   3077         self._config(config, self.func),
   3078         **kwargs,
   3079     )
   3080 else:
   3081     raise TypeError(
   3082         "Cannot invoke a coroutine function synchronously."
   3083         "Use `ainvoke` instead."
   3084     )

File ~/workplace/langchain/libs/core/langchain_core/runnables/base.py:975, in Runnable._call_with_config(self, func, input, config, run_type, **kwargs)
    971     context = copy_context()
    972     context.run(var_child_runnable_config.set, child_config)
    973     output = cast(
    974         Output,
--> 975         context.run(
    976             call_func_with_variable_args,
    977             func,  # type: ignore[arg-type]
    978             input,  # type: ignore[arg-type]
    979             config,
    980             run_manager,
    981             **kwargs,
    982         ),
    983     )
    984 except BaseException as e:
    985     run_manager.on_chain_error(e)

File ~/workplace/langchain/libs/core/langchain_core/runnables/config.py:326, in call_func_with_variable_args(func, input, config, run_manager, **kwargs)
    324 if run_manager is not None and accepts_run_manager(func):
    325     kwargs["run_manager"] = run_manager
--> 326 return func(input, **kwargs)

File ~/workplace/langchain/libs/core/langchain_core/runnables/base.py:2950, in RunnableLambda._invoke(self, input, run_manager, config, **kwargs)
   2948                 output = chunk
   2949 else:
-> 2950     output = call_func_with_variable_args(
   2951         self.func, input, config, run_manager, **kwargs
   2952     )
   2953 # If the output is a runnable, invoke it
   2954 if isinstance(output, Runnable):

File ~/workplace/langchain/libs/core/langchain_core/runnables/config.py:326, in call_func_with_variable_args(func, input, config, run_manager, **kwargs)
    324 if run_manager is not None and accepts_run_manager(func):
    325     kwargs["run_manager"] = run_manager
--> 326 return func(input, **kwargs)

Cell In[7], line 14, in call_tool(state)
     12 response = input(prompt=f"[y/n] continue with: {action}?")
     13 if response == "n":
---> 14     raise ValueError
     15 # We call the tool_executor and get back a response
     16 response = tool_executor.invoke(action)

ValueError: 

In [ ]: