CCAR-F exam topics
The CCAR-F exam (Claude Certified Architect (Foundations)) covers 5 domains and 30 topics. This page lists all of them with the weight of each domain, so you can see what the exam tests and where most of the marks are.
Domains and weights
Domain 1: Agentic architecture & orchestration · 27%
- Agentic Loops
On its own, Claude answers one request with one response. To act, such as looking up an order, it needs tools: functions in your code that you describe to Claude in the request. Claude does not run a tool you define. When it needs one, its response asks for the call, and your code runs the function and sends the result back. An agent is Claude, its tools and the code that keeps this exchange going.
- Multi-Agent Orchestration
Multi-agent orchestration is several Claude agents working together on one complex task. A coordinator agent receives the task and splits it up. Subagents each handle one specialised task. In the hub-and-spoke pattern, the coordinator is the hub and the subagents are the spokes. The coordinator's split of a task into subtopics is its decomposition.
- Subagent Invocation and Context Passing
A coordinator is the agent that runs a multi-agent task and hands parts of it to subagents. Invocation is how it starts a subagent. Context passing is how it gives that subagent the information it needs.
- Workflow Enforcement and Handoff
Some agent workflows have steps that must run in a set order. Prompt-based guidance sets that order with a system prompt instruction. Programmatic enforcement sets it in code: hooks, prerequisite gates or code-level checks that physically block downstream tools until their prerequisites complete.
- Agent SDK Hooks
An Agent SDK hook is code that intercepts a tool call or a tool result. Hooks are deterministic; the model is probabilistic. Hooks sit at the boundary between the model's decisions and the real world.
- Task Decomposition Strategies
Task decomposition breaks complex work into pieces an agentic system can handle. Fixed Sequential Pipelines (Prompt Chaining): Each step takes the previous step's output as input. The sequence does not change based on intermediate results.
- Session State and Resumption
A session is one conversation with an agent that you can come back to. Session management determines how an agent keeps continuity across work sessions.
Domain 2: Tool design & MCP integration · 18%
- Tool Interface Design
Tool descriptions are the primary mechanism LLMs use for tool selection. Minimal descriptions leave the model no way to tell overlapping tools apart.
- Structured Error Responses
A structured error response tells the agent the error's category, whether a retry can work, and how to recover. The MCP protocol provides the isError flag.
- Tool Distribution & Tool Choice
Tool distribution is the choice of which agent in a multi-agent system holds which tools. The number of tools an agent holds directly affects selection reliability.
- MCP Server Integration
An MCP server extends Claude by connecting it to an external system (database, API, development tool, issue tracker).
- Built-in Tools
Claude Code provides six built-in tools: Read, Write, Edit, Bash, Grep, Glob. Grep vs Glob: Grep, Glob; Grep finds what is inside files. Glob finds files by their names.
Domain 3: Claude Code configuration & workflows · 20%
- CLAUDE.md Hierarchy, Scoping, and Modular Organisation
A CLAUDE.md file is a Markdown file of instructions for Claude Code. Claude Code reads these files at three levels.
- Custom Slash Commands and Skills
A skill is a named, reusable set of instructions that Claude Code loads on demand. You run it as a slash command (/deploy).
- Path-Specific Rules for Conditional Convention Loading
Path-specific rules load conventions only when you edit files matching their glob patterns. They solve what neither root CLAUDE.md nor directory-level CLAUDE.md handles well: one file type scattered across many directories.
- Plan Mode vs Direct Execution
Claude Code has two main modes:; Plan mode: reads the codebase, analyses dependencies, proposes an approach — does not modify files until you approve.; Direct execution: makes changes straight away.
- Iterative Refinement Techniques
Iterative refinement steers Claude Code towards the right result over several rounds. Five techniques.
- CI/CD Integration
In CI/CD, Claude Code runs headless (no keyboard). Key patterns.
Domain 4: Prompt engineering & structured output · 20%
- System Prompts with Explicit Criteria
Vague instructions ("be conservative", "only report high-confidence findings") give the model no actionable decision boundary. Explicit categorical criteria define precisely what to flag and what to skip.
- Few-Shot Prompting
Few-shot prompting puts a few worked examples (input + desired output + reasoning) in the prompt. It is the most effective technique for consistent, well-formatted output.
- Structured Output with Tool Use
For guaranteed schema-compliant output: tooluse with JSON schemas (eliminates syntax errors) prompt-based JSON (no structural guarantees).
- Validation, Retry, and Feedback Loops
Schema syntax errors (malformed JSON, wrong types) are eliminated by tooluse + schemas. Semantic validation errors (sums don’t match, values in wrong fields) need external validation logic and retry loops.
- Batch Processing Strategies
Message Batches API: 50% cost savings, up to 24-hour window, no latency SLA, no multi-turn client tool calling within a single batch request, customid for correlation.
- Multi-Instance and Multi-Pass Review
Same-session self-review retains the original reasoning chain and tends to confirm its own decisions. Use an independent instance (fresh session, no prior context) for review.
Domain 5: Context management & reliability · 15%
- Context Window Management
The context window is everything Claude reads on one request. Managing it is choosing what goes in.
- Escalation & Ambiguity Resolution
When a tool returns multiple matches: always ask for additional identifiers. Never pick by "most recent" or "most active" heuristic (privacy risk).
- Error Propagation in Multi-Agent Systems
Middle ground: structured error propagation → coordinator decides recovery. Access Failure vs Valid Empty Result: Meaning, Retry?
- Codebase Exploration & Context Degradation
Context degradation = model loses grip on earlier specific findings (refers to "typical patterns" instead of class names). It is an attention quality problem, not a token-limit problem. Larger context windows do not fix it.
- Human Review & Confidence Calibration
97% overall accuracy can hide catastrophic failure rates on specific document types or fields (e.g. handwritten receipts at 60%). Always validate by document type and field segment before automating.
- Information Provenance & Multi-Source Synthesis
Every finding must include:; Claim; Source URL; Document name; Relevant excerpt; Publication date.
CCAR-F exam facts
| Questions | 60 |
|---|---|
| Time limit | 120 minutes |
| Question format | Multiple-choice and multiple-response, built around 4 scenarios drawn from a bank of 6 |
| Passing score | 720 (scaled) on a scale of 100–1,000 |
| Exam fee | $125 USD per attempt |
| Certification valid for | 12 months |
| Delivery | Pearson VUE, online proctored or at a test centre |
| Retakes | Waiting periods of 14, 30 and 90 days; at most 4 attempts in a rolling 12 months |
Study these topics
- Read the free CCAR-F study guide: one lesson per topic above, each with its sources in Anthropic's documentation.
- See how to prepare for CCAR-F for a study plan ordered by domain weight.
- Test yourself with the free 10-question CCAR-F mock test, or the full CCAR-F practice exam.
Skillmarked is independent and not affiliated with Anthropic. The topic list follows the public exam outline; check the official exam page for the current version before you book.