CLAUDE CODE AGENTS: ADVANCED ORCHESTRATION - Chapter 9, Exercise 2
A Task Run Twice a Year: Is a Pipeline Worth Building?
====================================================================================

QUESTION:
A developer is deciding whether to build a full multi-agent pipeline for a
task they expect to run roughly twice a year. Using this chapter's own tip
box, explain what they should conclude and why.

SOLUTION / EXPLANATION:

This chapter's tip box gives a direct method for this exact decision:
estimate how many times the workflow will genuinely be repeated before
committing to building a pipeline, since a pipeline's fixed setup cost only
pays off when it's amortized across real, repeated use. If the honest
estimate is "probably once or twice," the guidance is that a direct
approach or a single well-briefed agent is very likely more efficient
overall than the fixed cost of building a full pipeline.

Running roughly twice a year falls squarely into that low-repetition
category. Building several agent definitions, designing handoffs between
them, and setting up shared state (the real setup costs this course's own
earlier chapters described) all represent real, non-trivial upfront work.
Spread across only two uses per year, each individual use would have to
absorb roughly half of that entire setup cost just to break even against
simply doing the task directly (or with one single agent) each of those
two times - and that's before accounting for the ongoing cost of
maintaining a multi-agent pipeline that mostly sits unused between its
rare, twice-yearly runs.

The developer should conclude that a full pipeline is very likely not
worth building for this specific task. A direct approach, or at most a
single well-designed agent handling the whole task, is the more efficient
choice given how rarely this particular workflow actually recurs - the
pipeline's own upfront cost has too little repeated use to actually pay
itself back.

--------------------------------------------------------------------------
WHY THIS WORKS AS AN ANSWER:
It applies the chapter's own estimate-repetition-first method directly to
the twice-a-year frequency given, explains concretely why that frequency
is too low to amortize a pipeline's real setup cost, and reaches the
chapter's own recommended conclusion (favor the simpler approach) rather
than treating pipeline-building as automatically worthwhile.