Skip to main content
In this guide, we’ll walk through creating a custom LangChain tool that retrieves flight status details. You’ll learn how to define the tool, inspect its metadata, and wire it into a prompt–LLM chain for concise, one-word or short answers.

​
Defining the Tool

First, install or import the required packages: 
pip install langchain openai
Then define your tool using the @tool decorator from LangChain: 
from langchain.tools import tool
from langchain.llms.openai import OpenAI

@tool
def GetFlightStatus(flight_no: str) -> str:
    """Gets flight status and schedule"""
    # <Callout icon="triangle-alert" color="#FF6B6B">
# In production, replace this stub with a real API call.
    #
</Callout>
    return (
        f"Flight {flight_no} departed at 5:20 PM. "
        "It is on-time and expected to arrive at 8:10 PM at Gate B12."
    )
The @tool decorator registers the function’s name, description, and argument schema automatically.

​
Inspecting Tool Metadata

After defining GetFlightStatus, you can verify its registered metadata: 
print(GetFlightStatus.name)         # -> GetFlightStatus
print(GetFlightStatus.description)  # -> Gets flight status and schedule
print(GetFlightStatus.args)         
# -> {'flight_no': {'title': 'Flight No', 'type': 'string'}}
You can also view the complete StructuredTool representation: 
from langchain.tools import StructuredTool

print(StructuredTool(
    name='GetFlightStatus',
    description="Gets flight status and schedule",
    args_schema=GetFlightStatus.args_schema,
    func=GetFlightStatus.func
))
For a quick overview, here’s how the metadata maps out:
PropertyValue
nameGetFlightStatus
descriptionGets flight status and schedule
argumentsflight_no: string
return typestr

​
Using the Tool in a Chain

Next, we’ll connect the tool to a prompt template, the OpenAI LLM, and a simple output parser: 
from langchain import PromptTemplate
from langchain.output_parsers import StrOutputParser

prompt = PromptTemplate.from_template(
    "Based on the context: {context}\nAnswer the query: {query}"
)
llm = OpenAI()
output_parser = StrOutputParser()
Run the tool to generate the context: 
flight = "EK524"
context = GetFlightStatus.run(flight)
Compose the chain by piping the prompt into the LLM and then into the parser: 
chain = prompt | llm | output_parser

​
Invoking the Chain

With the chain ready, invoke it to extract specific details: 
print(chain.invoke({"context": context, "query": "status"}))         # -> On-time
print(chain.invoke({"context": context, "query": "departure time"})) # -> 5:20 PM
print(chain.invoke({"context": context, "query": "arrival time"}))   # -> 8:10 PM
print(chain.invoke({"context": context, "query": "gate"}))           # -> B12
Each invocation returns a concise answer tailored by your prompt design.
You now have a reusable flight status tool. Integrate real-world APIs inside GetFlightStatus to fetch live data, and combine multiple tools to build sophisticated LangChain agents.

Watch Video

Practice Lab