I couldn’t find any decent free practice material for the Claude Certified Architect (CCA) Foundations exam, so I built one.
The exam came out recently, and most of what’s available right now is either locked behind a paywall or too surface-level to be useful. I wanted something that actually reflects how the exam tests you — not definitions, but decisions.
So I put together a 100-question mock exam based entirely on scenario-driven problems.
Each question forces you to think through trade-offs: when to rely on prompt instructions vs programmatic enforcement, how to structure agent workflows, how to handle context and reliability, and so on. Basically the kind of judgment calls you’d make in a real system, not something you can memorize.
What it includes:
- 100 scenario-based questions across all 5 domains (same distribution as the real exam)
- Detailed explanations for every answer, including why the wrong options fail
- Ability to practice by domain, difficulty, or full mock sessions
- Works completely offline after first load (PWA)
- No login, no API key, no paywall
Domain split:
- Agentic Architecture & Orchestration — 27
- Claude Code Config & Workflows — 20
- Prompt Engineering & Structured Output — 20
- Tool Design & MCP Integration — 18
- Context Management & Reliability — 15
Live: https://neerajkr7.github.io/cca-foundations-exam-practice/
Repo: https://github.com/Neerajkr7/cca-foundations-exam-practice
It’s MIT licensed and open source.
One thing that stood out while building this: the hardest part of the exam isn’t syntax or APIs — it’s knowing when prompts are enough and when you need strict enforcement through tools, validation layers, or orchestration logic. That’s where most of the questions focus.
If you’re preparing, use it properly. Don’t just check answers — spend time understanding why you got something wrong. That’s where the actual learning happens.
And if you’ve already taken the exam, feel free to challenge the questions. If something’s off, open an issue or PR. I’d rather fix it than leave it misleading.

Top comments (0)