Free & open-source, no account required

CCA-F prep + deep AI engineering courses

Proctored mock exams, real activity tracking, and evidence-based recommendations for the CCA-F certification. Plus in-depth courses on Claude, agentic engineering,  OpenCode, loop engineering, and more.

Start CCA-F prepBrowse all courses

No card· No account· No limits

163

Lessons

Across CCA-F domains & courses

800+

Questions

Domain, mixed, and full mock formats

9

Courses

From Claude basics to loop engineering

720/1000

Pass score

Same scale as the real CCA-F exam

All courses

9 self-contained courses spanning 97 lessons — from Claude fundamentals to advanced loop engineering. Pick any path, learn at your pace.

Claude 101: Beginner's Guide
Learn Claude AI from the ground up, what Claude is, how to write effective prompts, projects, artifacts, skills, connectors, research mode, and more.

13 lessons

AI Fluency Foundations
Master the 4D Framework (Delegation, Description, Discernment, and Diligence) and build lasting skills for effective, efficient, and responsible human-AI collaboration.

13 lessons

Practical Exercises
Hands-on exercises for building production Claude applications: agent loops, Claude Code configuration, data extraction pipelines, and multi-agent research systems.

4 lessons

Latest Updates
Breaking developments in the Claude ecosystem, new model releases, policy changes, and frontier AI research.

1 lessons

Mastering Agentic Engineering
A deep, book-grounded journey from single-agent fundamentals through multi-agent orchestration, governance, and production deployment. Covers the full spectrum: agent architecture, perception, reasoning, action, reflection, multi-agent patterns, and production engineering.

12 lessons

AI Model Landscape
Compare Claude with GPT-4o, Gemini 2.5, DeepSeek, and GLM across architecture, API design, context limits, and use-case fit. Helps CCA-F candidates understand where Claude sits in the broader LLM ecosystem.

4 lessons

Claude for Your Team
Non-technical guide to understanding and using Claude — for PMs, designers, marketers, and writers working alongside AI-powered engineering teams.

8 lessons

OpenCode: The Open-Source AI Coding Agent
Master the open-source terminal-native AI coding agent. Learn agents, tools, configuration, permissions, skills, plugins, MCP integration, and CLI workflows from the source code up.

10 lessons

Loop Engineering: From Prompt to Autonomous Systems
Master the four layers of AI engineering — prompt engineering, context engineering, harness engineering, and loop engineering. Build reliable agent harnesses, design autonomous agent loops, and become a loop engineer.

8 lessons

The CCA-F exam, at a glance

A complete blueprint of what's tested, how it's weighted, and how to prioritize your study time. 60 questions · 120 minutes · 720/1000

Total questions60
Time limit120 min
Pass threshold720/1000
Question typeSingle-answer
Agentic Architecture27%
16 of 60 questions
Agent loopsMulti-agentPlanningSDK

The highest-weighted domain. Agent loops, planning, tool orchestration, multi-agent coordination, and the Claude Agent SDK (state, durable workflows, queues, retries, observability).

Claude Code20%
12 of 60 questions
CLICLAUDE.mdPlan modeHooks

Configure and drive Claude Code: CLAUDE.md, custom commands, plan mode, hooks, MCP integration, CI/CD, and team workflow patterns.

Prompt & Structured Output20%
12 of 60 questions
Few-shotCoTXMLJSON schema

Explicit criteria, few-shot, chain-of-thought, XML structuring, and reliable structured output via JSON schemas, tool_use, and validation-retry loops.

Tools & MCP18%
11 of 60 questions
Tool designMCP protoTransportsSecurity

Design effective tools (clear descriptions, structured errors, safe execution) and integrate the Model Context Protocol, architecture, resources, tools, prompts, transports, and security.

Context & Reliability15%
9 of 60 questions
Context mgmtCachingMonitoringRetries

Manage the context window (budgeting, compression, caching, RAG) and build for reliability: retries, fallbacks, guardrails, validation pipelines, and production hardening.

100%
0
720/1000 is the pass mark — shown as the amber bar above

Why your scores here are real

Three things that make this different from other CCA-F prep sites.

Proctored exam conditions
Mocks detect tab-switches, copy, DevTools, and fullscreen exits. You can't game your score, which means your mock score actually predicts your real one.
Real captured data
Every scroll, answer, and confidence rating is recorded locally. Your dashboard and readiness reflect what you actually did, never placeholder numbers.
Evidence-driven recommendations
Recency-weighted mastery, overconfidence detection, and scenario-vs-recall gaps rank what to study next, with a direct link to fix each gap.

Know where you are, every step

Courses for depth, CCA-F track for certification. Each phase feeds the next.

01your pace

Courses

9 courses from Claude 101 to loop engineering. Pick your path, learn in depth.

Browse courses
02~15 min

Diagnose

60-question diagnostic maps your CCA-F strengths and gaps across all 5 domains with an estimated score.

Take diagnostic
03daily

Learn & Practice

CCA-F lessons in domain-weight order, flashcards, mixed practice, and spaced repetition.

Open the path
04120 min

Mock & ready

Full-length mocks under proctoring. Readiness view tells you when to book the real exam.

Try a mock

Straight answers

Questions you're probably asking right now.

Start anywhere. Go deep.

Whether you're aiming for the CCA-F certification or mastering a specific skill — courses, mocks, and real data guide every step.

Start CCA-F prepBrowse all coursesTry a mock exam
FrontendX: not affiliated with AnthropicChangelogGitHubDownload book