Why observability is critical for agents in Agentic AI - Test Your Understanding

Practice - 5 Tasks

1fill in blank

easy

Complete the code to print the agent's current state for observability.

Agentic AI

print('Agent state:', [1])

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

Aagent.state

Bagent.run()

Cagent.execute()

Dagent.action

Attempts:

3 left

💡 Hint

Common Mistakes

Using methods like run() or execute() instead of accessing the state.

The agent's current state is stored in agent.state. Printing it helps observe what the agent is doing.

2fill in blank

medium

Complete the code to log the agent's decision for observability.

Agentic AI

log_entry = {'decision': [1]
logger.log(log_entry)

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

Aagent.input

Bagent.state

Cagent.memory

Dagent.last_action

Attempts:

3 left

💡 Hint

Common Mistakes

Logging the entire state or input instead of the last action.

The agent's last decision or action is stored in agent.last_action. Logging it helps track decisions.

3fill in blank

hard

Fix the error in the code to correctly track the agent's performance metric.

Agentic AI

performance = agent.metrics.get([1], 0)

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

A'accuracy'

Baccuracy

C'loss'

Dloss

Attempts:

3 left

💡 Hint

Common Mistakes

Forgetting quotes around dictionary keys.

Dictionary keys must be strings in quotes. Using 'accuracy' correctly accesses the metric.

4fill in blank

hard

Fill both blanks to create a dictionary comprehension that records agent states with timestamps.

Agentic AI

state_log = {timestamp: [1] for timestamp, [2] in agent.history.items()}

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

Astate

Caction

Devent

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'action' or 'event' instead of 'state' for the value.

The dictionary comprehension maps each timestamp to the corresponding state from the agent's history.

5fill in blank

hard

Fill all three blanks to filter and store agent actions with confidence above threshold.

Agentic AI

filtered_actions = [1]: [2] for [3], [2] in agent.actions.items() if [2]['confidence'] > 0.8}

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

Aaction_id

Baction

Attempts:

3 left

💡 Hint

Common Mistakes

Mixing up keys and values or variable names in the loop.

The comprehension uses action_id as keys and action as values, iterating over action_id, action pairs.