Introduction to Claude Code Subagents (course study guide)

A study guide to Anthropic Academy’s four-lesson Introduction to subagents course, written in this repository as Claude/subagents/README.md. It is an original summary, not a transcript.¹

Scope and coverage

Lesson	Coverage in the guide
What are subagents?	Article and video subtitles read in full
Creating a subagent	Article read in full; the video was blocked by a copyright claim on 2026-08-05, so no transcript was used
Designing effective subagents	Article and video subtitles read in full
Using subagents effectively	Article and video subtitles read in full

All course material was read on 2026-08-05.¹ Upstream sources are the four course articles on anthropic.skilljar.com and their YouTube videos; this wiki cites the study guide, not the course directly.

Takeaways

• A subagent trades visibility for a clean main context: its investigation stays isolated and only a summary returns.¹ See Subagent and Context isolation.
• The course calls a defined output format the most important design improvement, because it gives the subagent a stopping condition.¹ See Delegation contract.
• The single deciding question is whether the intermediate work matters to the main thread.¹ See When to delegate.
• Tools should be the minimum the role needs.¹ See Least-privilege tool access.
• Custom subagents are Markdown files with YAML frontmatter, created with /agents.¹ See Subagent configuration file.

Footnotes

1. Introduction to Claude Code Subagents (study guide) ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷