Unofficial exam prep · CCAR-F

Claude Certified Architect — Foundations (CCAR-F) practice questions

Prepare for the CCAR-F exam with practice questions mapped to all 5 domains of the published exam guide, flashcards and study decks — in English or Hebrew. Practising is free; Pro adds full-length timed mock exams and per-domain readiness analytics.

Start CCAR-F practice See what is free

Independent study tool — not affiliated with or endorsed by Anthropic.

The CCAR-F exam at a glance

Items: 60
Duration: 120 minutes
Pass mark: 720 / 1000 (scaled)
Format: multiple-choice items built around real-world scenarios
Delivery: Pearson VUE
Valid for: 12 months

Figures from the published exam guide (v1.0, effective July 2026). Always confirm current details with the official certification program before you book.

Who it is for: Solution architects, 6+ mo building with Claude API / Agent SDK / Claude Code / MCP.

What the CCAR-F exam covers

The trainer tags every CCAR-F practice question to one of these domains, so you can see where you are strong and drill where you are not.

CCAR-F exam domains and their weights
Domain	Exam weight
Agentic Architecture & Orchestration	27%
Claude Code Configuration & Workflows	20%
Prompt Engineering & Structured Output	20%
Tool Design & MCP Integration	18%
Context Management & Reliability	15%

Try a sample CCAR-F question

You're building a customer-support agent that runs as a loop between Claude API calls and tool executions (database lookups, refund issuance, ticket creation). The loop must terminate cleanly when Claude has finished helping the customer, and continue when Claude needs to invoke another tool. Two of your engineers propose competing termination signals. You need to pick the one that works reliably across thousands of conversations per day in production. Which signal should the loop check after each Claude response to decide whether to continue the loop?

Check the `stop_reason` field on the response — continue the loop on `tool_use`, exit on `end_turn`, raise on `max_tokens` or `refusal`
Check whether the assistant's text response contains terminal phrases like 'done', 'finished', or 'all set' — and return as soon as any matches
Maintain an iteration counter capped at 20 — after 20 cycles, force-exit regardless of the conversation state
Compare assistant message token count to a learned threshold — short messages signal conclusion, long messages indicate more work

Answers are graded in the trainer, with an explanation for every option. Answer it in the trainer →

How the trainer helps you prepare

Domain-tagged practice — questions written from the public exam guide, graded server-side with an explanation for each option.
Flashcards and study decks for the vocabulary and patterns each domain leans on.
Full-length timed mock exams that follow this track's exam length and clock (Pro).
Readiness analytics per domain, with practice rotated toward your weakest areas (Pro).
English and Hebrew interfaces — switch at any time without losing progress.

Pricing

USD 0 Free practice — practice questions, flashcards and study decks.
USD 29 one-time — unlock the whole Architect family (CCAR-F and CCAR-P).
USD 59 one-time — Pro: lifetime access to every live track, timed mocks and analytics.

Start CCAR-F practice

CCAR-F FAQ

What does the CCAR-F exam cover?

The published exam guide for Claude Certified Architect — Foundations weights 5 domains: Agentic Architecture & Orchestration (27%), Claude Code Configuration & Workflows (20%), Prompt Engineering & Structured Output (20%), Tool Design & MCP Integration (18%), Context Management & Reliability (15%).

How is the CCAR-F exam structured?

60 multiple-choice items built around real-world scenarios, 120 minutes, delivered through Pearson VUE. The pass mark is a scaled 720 out of 1000, and the certification is valid for 12 months. Each sitting draws 4 of 6 published scenarios.

Who is the CCAR-F certification for?

Solution architects, 6+ mo building with Claude API / Agent SDK / Claude Code / MCP.

How much does the CCAR-F trainer cost?

Practice questions, flashcards and study decks are free. Unlocking the whole Architect family (CCAR-F and CCAR-P) is USD 29 one-time; Pro (lifetime, every live track) is USD 59 one-time and adds full-length timed mock exams, per-domain readiness analytics and live team quizzes.

Is this trainer official or affiliated with Anthropic?

No. Cert Trainer by OpsAgents AI is an independent, unofficial study tool. It is not affiliated with, endorsed by, or sponsored by Anthropic, and its practice items are written from the public exam guides, not taken from any real exam. Claude is a trademark of Anthropic, PBC.