Prompt composition and chaining in LangChain - Step-by-Step Execution

Concept Flow - Prompt composition and chaining

Define Base Prompts

↓

Compose Prompts Together

↓

Chain Prompts Sequentially

↓

Pass Output of One as Input to Next

↓

Get Final Combined Result

Start by creating simple prompts, then combine them into a chain where each prompt's output feeds the next, producing a final result.

Execution Sample

LangChain

from langchain import PromptTemplate, LLMChain

# Define two prompts
prompt1 = PromptTemplate(input_variables=["name"], template="Hello {name}!")
prompt2 = PromptTemplate(input_variables=["greeting"], template="{greeting} How can I help you?")

This code defines two simple prompts: one greets by name, the other adds a question using the greeting.

Execution Table

Step	Action	Input	Output	Next Input
1	Run prompt1 with name='Alice'	name='Alice'	Hello Alice!	greeting='Hello Alice!'
2	Run prompt2 with greeting='Hello Alice!'	greeting='Hello Alice!'	Hello Alice! How can I help you?	None
3	Chain complete	None	Hello Alice! How can I help you?	End

💡 All prompts executed in sequence; final output produced.

Variable Tracker

Variable	Start	After Step 1	After Step 2	Final
name	None	Alice	Alice	Alice
greeting	None	Hello Alice!	Hello Alice!	Hello Alice!
final_output	None	None	Hello Alice! How can I help you?	Hello Alice! How can I help you?

Key Moments - 2 Insights

Why does the output of the first prompt become the input for the second?

Because in chaining, the output of one prompt is passed as input to the next, as shown in execution_table step 1 and 2.

Can prompts be chained in any order?

No, the order matters because each prompt may depend on the previous output, as seen in the flow from step 1 to step 2.

Visual Quiz - 3 Questions

Look at the execution table, what is the output after step 1?

A Hello Alice!

B Hello Alice! How can I help you?

C Alice

D How can I help you?

💡 Hint

Check the 'Output' column in row for step 1.

At which step does the final combined result appear?

A Step 1

B Step 2

C Step 3

D None

💡 Hint

Look at the 'Output' and 'Action' columns for step 3.

If the first prompt output changed, how would it affect the chain?

A No effect on the final output

B The second prompt input and final output change

C Only the first prompt output changes

D Chain breaks and stops

💡 Hint

Refer to how output of step 1 is input to step 2 in execution_table.

Concept Snapshot

Prompt composition and chaining:
- Define simple prompts with variables
- Combine prompts sequentially
- Output of one prompt feeds next
- Final output is combined result
- Order matters for correct chaining

Full Transcript

Prompt composition and chaining in Langchain means creating small prompts and linking them so the output of one becomes the input of the next. This lets you build complex conversations or tasks step-by-step. For example, a greeting prompt outputs 'Hello Alice!' which then feeds into a second prompt that adds 'How can I help you?'. The chain runs in order, producing a final combined message. This method helps organize prompts clearly and reuse parts easily.