How can I enforce strict adherence to the mail_tool input format in my LangChain agent?

Problem Description:

I am developing a LangChain agent with a custom email tool (mail_tool) designed to follow a strict set of rules and input format for sending emails. However, the agent sometimes invents email addresses or sends incorrect content that does not respect these rules.

Tool Rules and Input Format:

• The tool should only be used when the user requests email sending with a provided email address and content.
• It must use only the user-provided email; no invented emails.
• The input must be a JSON string representing a dictionary in the following format:

JSONCopied
{
    'mail': 'user_email',  # The user's email address.
    'content': 'content_to_send'  # The information to be sent.
}

Relevant Code:

PYTHONCopied
import asyncio
from langchain_openai import ChatOpenAI
from langchain.tools import Tool
from langgraph.checkpoint.memory import MemorySaver
from langchain_core.messages import HumanMessage, AIMessage
from langgraph.graph import StateGraph, START, END, MessagesState
from langchain.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain.agents import create_openai_functions_agent, AgentExecutor
import json

# Model and tools configuration
model = ChatOpenAI(model="gpt-4", openai_api_key="KEY_HERE")

prompt = ChatPromptTemplate.from_messages([
    ("system", "You are a friendly assistant named Axel, and you will answer questions. You only provide answers of a maximum of 25 words. You respond in Spanish"),
    MessagesPlaceholder(variable_name="chat_history"),
    ("human", "{input}"),
    MessagesPlaceholder(variable_name="agent_scratchpad")
])

def send_mail_tool(data: str):
    try:
        parsed_data = json.loads(data)
        mail = parsed_data.get("mail")
        content = parsed_data.get("content")
        if mail and content:
            return "Email sent"
        else:
            return "Error: 'mail' or 'content' is missing"
    except Exception as e:
        return f"Error: {str(e)}"

mail_tool = Tool(
    name="send_mail_tool",
    func=send_mail_tool,
    description="""
        FOLLOW THESE RULES STRICTLY TO USE THIS TOOL:
        - Use this tool only when the user requests to send information via email and provides their email address and the required content.
        - Only use it for this purpose.
        - Do not invent email addresses; only use the one provided by the user.
        - Always, without exception, use the Input format.
        - this is the Input Format:
            The input must be a dictionary with the following keys:
            {
                'mail': 'user_email',  # The user's email address.
                'content': 'content_to_send'  # The information to be sent.
            }
    """
)

tools = [mail_tool]

agent = create_openai_functions_agent(llm=model, prompt=prompt, tools=tools)
agent_executor = AgentExecutor(agent=agent, tools=tools, verbose=False)

workflow = StateGraph(state_schema=MessagesState)

async def call_agent(state: MessagesState):
    user_input = state["messages"][-1].content
    chat_history = state["messages"][:-1]
    response = await agent_executor.ainvoke({
        "input": user_input,
        "chat_history": chat_history
    })
    return {"messages": state["messages"] + [AIMessage(content=response["output"])]}

workflow.add_node("agent", call_agent)
workflow.add_edge(START, "agent")

memory = MemorySaver()
app = workflow.compile(checkpointer=memory)

config = {"configurable": {"thread_id": "user1234"}}

app.update_state(config, {"messages": [
    AIMessage(content="You are a friendly assistant named Axel, and you will answer questions. You only provide answers of a maximum of 25 words. You respond in Spanish")
]})

async def main():
    print("Assistant Online. Type 'exit' to quit.")
    while True:
        user_input = input("\nYou: ")
        if user_input.lower() == "exit":
            print("Goodbye!")
            break
        app.update_state(config, {"messages": [HumanMessage(content=user_input)]})
        state = app.get_state(config).values
        try:
            output = await app.ainvoke(state, config)
        except Exception as e:
            print("\nAssistant: Sorry, an error occurred while processing your request.")
            print(f"Details: {str(e)}")
            continue
        print("\nAssistant:")
        print(output["messages"][-1].content)

if __name__ == "__main__":
    asyncio.run(main())

Specific Issue:

• The agent sometimes invents email addresses instead of using the provided one.
• It may fail to send the complete content or adhere to the specified input format.

Question:
How do I modify the tool or agent so that it strictly follows the email sending rules and input format? Are there alternative approaches to enforce consistency, such as additional validations or using structured tools?

Additional Information:

• I am using GPT‑4 as the language model.
• The send_mail_tool function simulates sending an email and must adhere exactly to the provided format and rules.

I would appreciate any suggestions to achieve a more reliable behavior from the agent. Thank you!