A scenario-based practice exam built from a 90-question bank covering agentic architecture, MCP and tool design, Claude Code workflows, prompt engineering, and context reliability. Answer choices are shuffled on every run, so you learn the reasoning, not the position.
Sit a 60-question mock across four domain-balanced scenarios, or drill a single scenario or a quick 20-question set.
Answers never leave the server, so the score is computed centrally and can't be faked. A full per-option review is revealed once you submit.
Get a 100 to 1000 scaled score, a pass or fail verdict, per-scenario breakdown, and a full answer review.
This is an independent study and revision tool for exam preparation. It is not affiliated with, endorsed by, sponsored by, or connected to Anthropic. "Claude", "Claude Code", and related product names are referenced for descriptive study purposes only and remain the property of their respective owners. The visual styling is an original interpretation; no proprietary logos, fonts, or assets are used. Question content is supplied by the exam administrator for their own students.
We record your name and email so your result can be attributed to you. Nothing else is collected.
Four curated scenarios covering all five domains, 15 questions each. 120-minute limit, exam-style.
Focus on one scenario at a time. Untimed, great for targeted revision.
Twenty random questions across all scenarios. Untimed warm-up.
Your answers are scored on the server and your result is recorded for the exam administrator. If submission fails, your answers are kept in this browser so you can retry.