CLAUDE CODE AGENTS: FUNDAMENTALS - Chapter 3, Exercise 2
Foreground or Background When the Very Next Step Depends on the Result
====================================================================================

QUESTION:
You need a research agent's findings before you can decide your very next
step in a task. Explain whether you should launch it in the foreground or
the background, and why, using this chapter's own comparison.

SOLUTION / EXPLANATION:

This chapter's own comparison is direct on exactly this situation:
foreground is the right choice specifically when you genuinely need the
agent's result before you can take your very next step - which is exactly
what's described here.

Launching this research agent in the background would let you keep working
on something else while it runs, but that only helps if there's a
meaningful, independent next step available to work on in the meantime.
Since the scenario states the very next step in this task depends entirely
on what the research agent finds, there's nothing productive to do on this
particular task until that result comes back - continuing to talk or act as
though you already know the answer, before the agent has actually reported
back, is precisely the mistake this chapter warns against under "Don't Race
Ahead of a Background Agent."

Running the agent in the foreground instead means the conversation
naturally pauses until the agent finishes and returns its findings
immediately - at which point you have the actual information needed to
decide the next step, with no risk of guessing or proceeding on an
assumption instead of a real result.

So the correct choice is foreground, specifically because this chapter
frames that choice around whether the result is needed before your very
next step - not around how long the task might take or how important it
is in the abstract.

--------------------------------------------------------------------------
WHY THIS WORKS AS AN ANSWER:
It applies the chapter's own stated criterion (whether the result is
needed before the next step) directly to the scenario, rather than
reasoning from task importance or duration, and explicitly connects the
background alternative to the specific risk (racing ahead of results) the
chapter names elsewhere.