Claude Code · Free micro-course

Stop burning Claude Code's context on reads you'll never look at twice.

In 30 minutes, you'll know when to delegate work to a subagent, and when delegating is the worst thing you can do.

Taught by Edwin Ashdown. Founder of AIFirstPartner, 12 production AI systems shipped. Also builder of AI IQ and TranscriptIQ.

You're ten minutes into a Claude Code session and you ask a simple question: which service handles refunds? Claude reads fifteen files, runs a handful of searches, traces a few function calls, and answers. The answer is correct. But every file it touched is now sitting in your main context, and you only needed one fact. Subagents fix that. You'll build your first one in lesson 2, and by lesson 4 you'll have a single-question rule for when to use them.

Start lesson 1 (4 min read)

What you'll be able to do

Ship a code-review subagent that catches issues your main thread misses
Cut a long Claude Code session's context usage by delegating research that was clogging your window
Know in one question whether a task belongs in a subagent or the main thread
Write subagent descriptions that trigger reliably instead of sitting unused

What you'll walk away with

Subagents, end to end

4 lessons· Free

Why one simple question filled your context with 15 files
Build your first custom subagent in two minutes
Four patterns that stop subagents wandering
When to delegate, when not to, and the one question that decides

Open lesson 1 →

About the instructor

Edwin Ashdown founded AIFirstPartner and has shipped 12 production AI systems for clients in government, legal, and enterprise. He is fractional CTO at Speculo, a cyber risk software company, and manages cloud and data platform engineering for a large Australian government department supporting 20+ agencies across the portfolio.

He publishes his build process at 150days.aifirstpartner.com.

This course is the playbook he uses when onboarding a team onto Claude Code subagents.

Frequently asked questions

Why not just read Anthropic's subagent docs?

The docs tell you what subagents are. This course tells you which ones to skip (expert personas, sequential pipelines, test runners) and gives you a seven-section output template you can paste into any config. Thirty minutes here versus a weekend of trial and error.

Is this affiliated with Anthropic?

No. Introduction to Subagents is independent. This course is not an Anthropic product and has not been reviewed or endorsed by them.

Who is this course for?

If you've used Claude Code enough to hit a context-full warning, you're ready. If you've never opened it, start with Claude Code 101 first. If you've already shipped three custom subagents, the decision framework in lesson 4 is worth 5 minutes; skip the earlier lessons.

Is this course free?

Yes. All four lessons are free and immediately accessible. No account or payment required.

How long does the course take?

About 30 minutes. Four focused lessons, each covering a specific aspect of subagents in Claude Code.

Do I need prior experience with subagents?

No. The course starts with what subagents are and builds up to creating, designing, and deploying them effectively. Basic familiarity with Claude Code is the only prerequisite.

Will I build something during the course?

Yes. The course walks through creating a code-quality-reviewer subagent as the running example, with the full config file in lesson 2. You can follow along in Claude Code and have your first custom subagent committed to .claude/agents/ by the end of lesson 2.

The next time your main context fills up, you'll know what to do about it.

Four lessons. About 30 minutes. Free.

Start lesson 1 (4 min read)