Prompt composition and chaining in LangChain - Mini Project: Build & Apply

Building a Simple Prompt Chain with Langchain

📖 Scenario: You are creating a small program that uses Langchain to combine two prompts. The first prompt asks for a user's favorite color. The second prompt uses that answer to create a sentence describing the color.

🎯 Goal: Build a Langchain prompt chain that first asks for a favorite color, then uses that answer to generate a descriptive sentence.

📋 What You'll Learn

Create a Langchain prompt template for asking the favorite color

Create a second prompt template that uses the color answer

Chain the two prompts so the output of the first is input to the second

Run the chain and get the final output

💡 Why This Matters

🌍 Real World

Prompt chaining is useful when you want to break down a complex question into smaller steps and combine their answers smoothly.

💼 Career

Understanding prompt composition and chaining is important for building conversational AI, chatbots, and automated assistants that handle multi-step interactions.

1

Create the first prompt template

Create a Langchain prompt template called color_prompt with the template string: 'What is your favorite color?'

LangChain

# Create the first prompt template called color_prompt
# Your code here

Need a hint?

Use PromptTemplate from langchain.prompts and assign it to color_prompt.

2

Create the second prompt template using the color

Create a Langchain prompt template called description_prompt with the template string: 'The color you chose is {color}.' and specify input_variables=['color']

LangChain

from langchain.prompts import PromptTemplate

color_prompt = PromptTemplate(template='What is your favorite color?')

# Create the second prompt template called description_prompt
# Your code here

Need a hint?

Remember to include input_variables=['color'] so the prompt knows to expect a variable named color.

3

Chain the two prompts using SimpleSequentialChain

Import SimpleSequentialChain from langchain.chains and create a chain called color_chain that runs color_prompt then description_prompt in sequence

LangChain

from langchain.prompts import PromptTemplate

color_prompt = PromptTemplate(template='What is your favorite color?')

description_prompt = PromptTemplate(template='The color you chose is {color}.', input_variables=['color'])

# Import SimpleSequentialChain and create color_chain
# Your code here

Need a hint?

Use SimpleSequentialChain and pass a list of the two prompt templates in order.

4

Run the chain and get the final output

Call color_chain.run() and assign the result to a variable called final_output

LangChain

from langchain.prompts import PromptTemplate
from langchain.chains import SimpleSequentialChain

color_prompt = PromptTemplate(template='What is your favorite color?')

description_prompt = PromptTemplate(template='The color you chose is {color}.', input_variables=['color'])

color_chain = SimpleSequentialChain(chains=[color_prompt, description_prompt])

# Run the chain and assign the output to final_output
# Your code here

Need a hint?

Use the run() method on color_chain and save the result in final_output.