4 lessons, each a self-contained pack — study notes, flashcards, and a spoken recap. Open offline; learn at your own pace.
Each card opens a full lesson: notes to read, cards to drill, and a recap to listen to. Start anywhere; they stand alone.
Subagents are specialized assistants that Claude Code can delegate tasks to. Think of them as focused helpers: each one runs in its own conversation context window, does its work, and returns a summary to the main thread. The intermediate steps -- all the file reads, searches, and tool calls -- stay isolated and never clutter your main conversation.
Claude Code comes with built-in subagents, but you can also create your own. Custom subagents specialize in specific tasks -- like reviewing code, writing tests, or checking documentation. They are defined as markdown files with YAML frontmatter that tell Claude when to use the subagent and how the subagent should behave.
Now that you know how to create subagents, let's look at the patterns that make them actually effective. A subagent that's poorly configured will wander, run too long, or produce output the main agent can't use. The fixes come down to four things: writing good descriptions, defining an output format, reporting obstacles, and limiting tool access.
You know how to create subagents and design them well. Now the question is: when do they actually help, and when do they get in the way? The difference comes down to one thing -- whether the intermediate work matters to your main thread.