Exam-realistic questions
Every attempt generates 20 fresh questions across the full Foundations blueprint, timed like the real thing.
Claude Certified Developer
Walk into the Foundations exam already knowing how it feels.
Timed mock exams, an explanation behind every answer, and a personal study plan built from the topics you actually miss.
Progress that follows you
Sign in with your email and your attempts, last score and running average are always waiting for you.
Study notes on your gaps
After each exam you get a topic breakdown, then generate focused notes for the topics that cost you points.
What gets tested
Every attempt mirrors the official blueprint: questions are allocated across the eight exam domains in the same proportion as the real certification.
- Applications and Integration33.1%
Messages API request/response shape and parameters · multi-turn conversations and message roles · streaming responses and event handling · multimodal input (images, PDFs, documents)
- Model Selection and Optimization16.8%
choosing between Claude model tiers for a workload · cost and latency trade-offs · token accounting and context window limits · prompt caching
- Agentic Architecture and Workflows14.7%
agent loops and tool-use cycles · orchestrator/sub-agent patterns · chaining, routing and parallelization workflows · memory and state across turns
- Prompt and Context Engineering11%
system prompts and role setting · clear instructions, examples and few-shot prompting · XML tags and output formatting · chain-of-thought and reasoning prompts
- Tools and Model Context Protocol (MCP)10.6%
tool definitions and JSON input schemas · tool_use and tool_result blocks · parallel tool calls and tool choice · MCP servers, clients, resources and prompts
- Security and Safety8.1%
prompt injection and untrusted content · API key handling and server-side secrets · PII handling and data minimization · guardrails, refusals and content moderation
- Claude Code3.1%
Claude Code CLI workflows · project context files and configuration · permissions and allowed tools · integrating Claude Code into development loops
- Evaluation, Testing, and Debugging2.6%
building evaluation sets and graders · measuring accuracy and regressions · debugging bad outputs and tool errors · logging, tracing and observability