LangChain agents overview in Agentic AI - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to create a basic LangChain agent using an LLM.

Agentic AI

from langchain.agents import initialize_agent
from langchain.llms import OpenAI

llm = OpenAI(temperature=0)
agent = initialize_agent(llm, tools=[], agent= [1] )

Drag options to blanks, or click blank then click option'

A"toolkit-agent"

B"chat-completion"

C"simple-agent"

D"zero-shot-react-description"

Attempts:

3 left

💡 Hint

Common Mistakes

Using an invalid or unsupported agent type string.

Passing the agent type without quotes.

The initialize_agent function requires the agent type as a string. "zero-shot-react-description" is a common agent type for LangChain agents that react to inputs and decide actions.

2fill in blank

medium

Complete the code to add a tool to the LangChain agent.

Agentic AI

from langchain.agents import Tool

tool = Tool(name="Search", func=lambda x: "Searching..." + x, description="Search tool")
agent = initialize_agent(llm, tools=[[1]], agent="zero-shot-react-description")

Drag options to blanks, or click blank then click option'

Atool

Btools

Csearch_tool

DSearch

Attempts:

3 left

💡 Hint

Common Mistakes

Passing the string name of the tool instead of the Tool object.

Passing an undefined variable.

The tools parameter expects a list of Tool objects. Here, the variable tool is the correct object to pass.

3fill in blank

hard

Fix the error in the code to run the LangChain agent on an input.

Agentic AI

response = agent.run([1])
print(response)

Drag options to blanks, or click blank then click option'

Ainput("What is the weather today?")

BWhat is the weather today?

C"What is the weather today?"

Dagent("What is the weather today?")

Attempts:

3 left

💡 Hint

Common Mistakes

Passing the question without quotes causing syntax errors.

Using input() function inside run() which is not needed here.

The run method expects a string input. The input must be a string literal or variable containing a string. Passing a raw string without quotes causes a syntax error.

4fill in blank

hard

Fill both blanks to create a custom tool and add it to the agent.

Agentic AI

def custom_func(input_text):
    return f"Processed: {input_text}"

custom_tool = Tool(name=[1], func=[2], description="Custom processing tool")
agent = initialize_agent(llm, tools=[custom_tool], agent="zero-shot-react-description")

Drag options to blanks, or click blank then click option'

A"CustomTool"

Bcustom_func

C"custom_func"

DcustomTool

Attempts:

3 left

💡 Hint

Common Mistakes

Passing the function name as a string instead of a function object.

Not quoting the tool name string.

The Tool constructor requires the name as a string and the function as a callable. So the name is a string like "CustomTool" and the function is the function object custom_func.

5fill in blank

hard

Fill all three blanks to create an agent that uses multiple tools and runs a query.

Agentic AI

tool1 = Tool(name=[1], func=lambda x: f"Tool1: {x}", description="First tool")
tool2 = Tool(name=[2], func=lambda x: f"Tool2: {x}", description="Second tool")
agent = initialize_agent(llm, tools=[tool1, tool2], agent="zero-shot-react-description")
response = agent.run([3])
print(response)

Drag options to blanks, or click blank then click option'

A"ToolOne"

B"ToolTwo"

C"How do I use these tools?"

D"Tool1"

Attempts:

3 left

💡 Hint

Common Mistakes

Using the same name for both tools causing confusion.

Passing the query without quotes.

The first tool is named "Tool1" (matching the variable), the second is "ToolTwo" as a distinct name, and the query passed to run must be a string question.