CLAUDE CODE AGENTS: ADVANCED ORCHESTRATION - Chapter 1, Exercise 1
Why Orchestration Needed a Whole Second Course
====================================================================================

QUESTION:
A colleague who finished Claude Code Agents: Fundamentals asks why
orchestration needed a whole second course rather than one more chapter
added to the first. Using this chapter's own tip box, explain the
reasoning.

SOLUTION / EXPLANATION:

This chapter's tip box is explicit that orchestration is a genuinely
different skill from single-agent design, not simply "more of the same
material." Fundamentals' entire eleven chapters focused on one question:
how do you design and brief ONE agent well, whether that's a teaching
agent, a coding agent, or any of the other five types covered. Every skill
in that course - writing a good description field, choosing the right
tools, briefing an agent with enough context - answers that one-agent
question.

Orchestration introduces problems that simply don't exist when only one
agent is involved at a time. A chain of agents has multiple handoff points,
each one a fresh opportunity for context to get lost unless it's
deliberately carried forward - Chapter 1's own "an agent starts cold"
problem doesn't just apply once in a chain, it applies at every single
handoff. Coordinating whether agents run in parallel or in sequence,
passing state between them, and debugging which specific agent in a chain
actually caused an incorrect result are all genuinely new concerns with no
real equivalent in single-agent work.

Given that, folding orchestration into one extra chapter of Fundamentals
would have either shortchanged these genuinely new problems into a
surface-level treatment, or bloated a course that was otherwise focused
entirely on single-agent design. Splitting it into its own course lets
orchestration get the depth this chapter argues it actually requires,
while keeping Fundamentals focused on the foundational skill it was always
about.

--------------------------------------------------------------------------
WHY THIS WORKS AS AN ANSWER:
It identifies specifically what kind of problem orchestration introduces
that single-agent design never encounters (multiple handoff points, each
its own opportunity for lost context), and explains why that genuine
difference in depth and subject matter justifies a full second course
rather than a single additional chapter.