Until now, aicert.study covered CCAR-F from one angle only: a large question bank and mock exams to test whether you already knew the material. That's useful for checking readiness, but it assumes you got there some other way — scattered docs, the official exam guide's bullet points, or trial and error in production.
The Study Track is the piece we were missing: a linear, free path through everything CCAR-F actually tests, with working code you run yourself instead of slides you read once.
What's in it
21 lessons plus a capstone, ordered the way the decisions actually build on each other — not the way a spec sheet lists them. Every lesson maps to one or more of CCAR-F's five official domains:
| Domain | Weight | What it covers |
|---|---|---|
| Agentic Architecture & Orchestration | 27% | Tool-calling loops, multi-agent delegation, when not to add another agent |
| Claude Code | 20% | Claude Code for dev teams, memory files, skills, CI integration |
| Prompt Engineering | 20% | Task decomposition, structured output, defensive parsing, reliable extraction |
| Tool Design & MCP Integration | 18% | MCP server design, tool contracts, progressive discovery, error-as-response |
| Context Management & Reliability | 15% | Caching, long-context provenance, evals, observability |
Each lesson follows the same shape: a concrete problem, the concept that resolves it, a decision framework you can apply beyond the example, a build it section with a small hands-on lab, and a 6-question checkpoint quiz (rendered inline, no separate page) before you move on.
The labs live in a companion repo, labs/ccar-f/lessons — clone it, run the code, break it on purpose. We didn't want a browser sandbox that hides what's actually happening; you get the real files and a terminal.
It tracks your progress
If you're logged in, the track remembers which lessons you've finished and which quiz you're on, so you can pick up where you left off instead of re-scanning a page of lesson titles every time.
Where it ends
The last lesson is a capstone: defend a full CCAR-F architecture across all five domains at once, the same way a real exam question forces you to reason across domains instead of one at a time.
Free, on purpose
The Study Track costs nothing and always will — it's how we want people to actually get good at this, not just pass a test. The paid mock exam is still there for the part the track doesn't replace: timed, scored practice under exam conditions across our full question bank.
Start the CCAR-F Study Track for free →
A note on where the structure came from
The lesson sequence, domain mapping, and problem → concept → decision → build-it shape are adapted from rohitg00/ai-engineering-from-scratch (MIT licensed) — a genuinely well-designed curriculum skeleton. Every word of lesson content, every quiz question, and every lab on aicert.study was written from scratch in our own voice; we only borrowed the pedagogical structure, not the prose. Full attribution is on our about page.
Free CCAR-F practice exam, covering all 5 domains: aicert.study/certifications/ccar-f.