Can I combine Agent and @tool and create a function for LangGraph?

[mraguth]

I am referencing this link : https://langchain-ai.github.io/langgraph/tutorials/multi_agent/multi-agent-collaboration/#define-tools

In this example 'chart_agent' is calling python_repl tool and performaing the action. However, I like to combine both tasks and create a function by that i can call in workflow.add_node. Is that possible? please help.

Tool

repl = PythonREPL()


@tool
def python_repl(
    code: Annotated[str, "The python code to execute to generate your chart."],
):
    """Use this to execute python code. If you want to see the output of a value,
    you should print it out with `print(...)`. This is visible to the user."""
    try:
        result = repl.run(code)
    except BaseException as e:
        return f"Failed to execute. Error: {repr(e)}"
    result_str = f"Successfully executed:\n\`\`\`python\n{code}\n\`\`\`\nStdout: {result}"
    return (
        result_str + "\n\nIf you have completed all tasks, respond with FINAL ANSWER."
    )

Agent:

# chart_generator
chart_agent = create_agent(
    llm,
    [python_repl],
    system_message="Any charts you display will be visible by the user.",
)
chart_node = functools.partial(agent_node, agent=chart_agent, name="chart_generator")

[dosubot[bot]]

Yes, you can combine the Agent and @tool to create a function for LangGraph that can be called in workflow.add_node. Here's how you can structure it:

1. Define the Tool: Use the @tool decorator to define your tool, which in this case is python_repl.

2. Create a Function: Define a new function that encapsulates the logic of calling the python_repl tool.

3. Add to Workflow: Use workflow.add_node to add this function to your workflow.

Here's an example of how you might implement this:

from langchain_core.tools import tool
from langgraph.graph import StateGraph
from langchain_core.agents import create_agent
from typing import Annotated

# Define the PythonREPL tool
repl = PythonREPL()

@tool
def python_repl(
    code: Annotated[str, "The python code to execute to generate your chart."],
):
    """Use this to execute python code. If you want to see the output of a value,
    you should print it out with `print(...)`. This is visible to the user."""
    try:
        result = repl.run(code)
    except BaseException as e:
        return f"Failed to execute. Error: {repr(e)}"
    result_str = f"Successfully executed:\n\`\`\`python\n{code}\n\`\`\`\nStdout: {result}"
    return (
        result_str + "\n\nIf you have completed all tasks, respond with FINAL ANSWER."
    )

# Create an agent using the tool
chart_agent = create_agent(
    llm,
    [python_repl],
    system_message="Any charts you display will be visible by the user.",
)

# Define a function that uses the agent
def chart_generator_function(state):
    # Logic to use the agent
    return {"messages": ["Chart generated successfully"]}

# Create a new graph
workflow = StateGraph(AgentState)

# Add the function as a node in the workflow
workflow.add_node("chart_generator_node", chart_generator_function)

# Set the entry point and define edges as needed
workflow.set_entry_point("chart_generator_node")

This setup allows you to encapsulate the logic of using the python_repl tool within a function and add it to your LangGraph workflow ^[1].

To continue talking to Dosu, mention @dosu.

[mraguth]

@dosu How I use the chart_agent here?

# Define a function that uses the agent
def chart_generator_function(state):
    # Logic to use the agent
    return {"messages": ["Chart generated successfully"]}

[dosubot[bot]]

To use the chart_agent within the chart_generator_function, you need to invoke the agent with the appropriate input and handle its output. Here's how you can modify the function to utilize the chart_agent:

def chart_generator_function(state):
    # Example input for the agent
    input_data = "Your input data or prompt here"

    # Use the chart_agent to process the input
    result = chart_agent(input_data)

    # Update the state with the result
    state['messages'].append(result)

    return state

In this example, replace "Your input data or prompt here" with the actual input you want to provide to the chart_agent. The result from the agent is appended to the messages in the state, which allows you to maintain the conversation state or accumulate results.

To continue talking to Dosu, mention @dosu.