CLAUDE CODE AGENTS: FUNDAMENTALS - Chapter 11 (Capstone), Exercise 3
What to Weigh Before Building All Seven Agents Upfront
====================================================================================

QUESTION:
A developer decides to build and deploy all seven of these agents before
generating even a single course chapter for this project. Using this
chapter's own warning box, explain what consideration they should weigh
first.

SOLUTION / EXPLANATION:

This chapter's warning box is explicit that assembling a full agent team
has a real setup cost of its own, and that this cost pays off specifically
at repeated volume - not for using agents as a goal in themselves, and not
necessarily for a single one-off task. Building and properly configuring
seven distinct agent definitions (each with its own description, tool
allowlist, system prompt, and model choice) before generating even one
chapter means paying that entire setup cost upfront, before any of it has
actually been put to use or proven worthwhile for this specific project's
real workflow.

The consideration worth weighing first is genuinely how much repeated,
similar work is actually expected. If this project only occasionally
generates a new course chapter, much of that setup effort may go
underused relative to the time it took to build - the same discipline
this chapter itself notes is often reasonably done by hand for a single
task, without the overhead of formalizing a full team first. If, on the
other hand, chapters are generated frequently and repeatedly, at real
volume, across many courses over time, the same fixed setup cost gets
amortized across many repeated uses, and the payoff described in this
chapter's tip box (chaining these steps automatically, per Claude Code
Agents: Advanced Orchestration) becomes genuinely worth the investment.

So the developer's real question isn't "would these seven agents be
useful in principle" - almost any well-designed agent could theoretically
help somewhere - but "is the actual expected volume and repetition of this
work high enough to justify paying the setup cost of all seven now, rather
than building them incrementally as the actual need for each one becomes
clear."

--------------------------------------------------------------------------
WHY THIS WORKS AS AN ANSWER:
It identifies the specific tradeoff the warning box names (setup cost vs.
payoff at volume) and translates it into a concrete question the developer
should actually ask (how much repeated work is genuinely expected) rather
than treating "build everything now" as either obviously right or
obviously wrong.