CCAR-F FAQ
Straight answers about Claude Certified Architect - Foundations: what it costs, how it's scored, how retakes and renewal work, and the concepts it tests. Exam logistics come from the sourced exam facts; the concept answers are rewritten from this site's own glossary.
How much does the CCAR-F exam cost?
Claude Certified Architect - Foundations costs $125 per attempt, as listed on Anthropic's Partner Academy certifications FAQ (before any Partner Network tier discount). A retake is the full exam fee, and your tier discount applies to retakes too.
How many questions are on the CCAR-F exam and how long is it?
60 questions in 120 minutes, a mix of multiple choice and scenario-based multiple response, delivered by Pearson VUE online or at a test center. The FAQ advises planning for about 135 minutes of total seat time.
What score do I need to pass CCAR-F?
A scaled score of 720 out of 1000. It is a scaled score, not a flat percent-correct cutoff, so the exact number of correct answers required is not published.
How long do I have to wait to retake CCAR-F after failing?
Waits grow with each failed attempt: 14 days, then 30 days, then 90 days after the first, second, and third failed attempts. You can sit the exam up to 4 times per rolling 12 months.
Are there prerequisites for CCAR-F?
No formal prerequisite - no Claude certification requires another. Registration does require a work email on a Claude Partner Network organization's domain.
How long is CCAR-F valid, and how do I renew it?
It is valid for 12 months from the date you earn it. Free, non-proctored reassessment if renewed on time. Let it lapse and you retake the full proctored exam at full cost. No grace period is stated in the Partner Academy FAQ. Treat the 12-month date as a hard deadline and renew before it - if it lapses, the FAQ says you retake the full exam at the full fee.
What does the CCAR-F exam cover?
Its published blueprint has 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%).
Who is CCAR-F designed for?
Practitioners and consultants who design agentic systems on Claude end to end.
How does an agentic loop know when to stop?
In an agentic loop Claude receives input, chooses an action (usually a tool call), observes the result, and repeats. The dependable stop signal is the response's stop_reason reaching end_turn. Checking the text for "done", or making an iteration cap the primary control, are the classic wrong answers - a cap is only a safety net.
Which orchestration patterns should I know for CCAR-F?
Four recur: prompt chaining (sequential calls), routing (directing input to a specialised handler), parallelisation (concurrent calls), and orchestrator-workers (a central agent delegating to sub-agents). The exam tests choosing between them for a described workload.
What does prompt caching do, and where does the cache breakpoint go?
It lets frequently reused prompt content be cached, cutting latency and cost on later requests, with cache hits billed at a reduced rate. Put the cache_control breakpoint at the end of the static block - system prompt, tool definitions, large documents - and order dynamic content after it. The default ephemeral cache lives for roughly five minutes.
What are hooks in Claude Code?
Custom scripts that run at set points in Claude Code's lifecycle - before or after tool calls, on notification events, or at session start. They are defined in settings.json and run as shell commands on your own machine.
How do Claude Code permissions work?
They use an allowlist/denylist model: you explicitly permit or deny specific tools, file paths, and commands. In the default mode Claude Code asks for confirmation before potentially destructive operations.
What is an MCP server, and how does it differ from an MCP client?
An MCP server exposes tools, resources, and prompts over the Model Context Protocol and handles incoming tool calls; it can run locally or remotely. The client connects to one or more servers, discovers their capabilities, and forwards Claude's tool calls to the right one. Claude Code has a client built in; in your own application you implement it.