Does this course prepare me for the CCA-F certification?

It is a strong hands-on supplement. The course gives you working depth in Claude Code configuration and workflows, tool design and MCP, agentic architecture, prompting, and context management, which together span all five CCA-F domains. Pair it with our CCA-F flashcards and practice tests for exam-format coverage of every tested detail.

I don't write code for a living. Can I still take it?

Levels 1 and 2 assume nothing beyond comfort in a terminal and serve anyone who works with Claude, including Cowork users. From Level 4 onward the material is developer-focused: you will be editing repositories, writing configuration, and eventually using the Agent SDK.

How is this different from Anthropic's official Claude Code course?

Anthropic's Claude Code in Action is a solid one-hour video primer, and we recommend it. This course is the graded, hands-on program that goes deeper: 8 levels of do-first challenges with instant grading, verify-it-yourself terminal steps, and rubric-graded boss projects, plus explicit CCA-F exam alignment.

Course·Beginner · 20–30 hours

Master Claude Code, level by level

Predict how Claude Code behaves, prove it in your own terminal, ship real work. Eight levels, from your first session to the Agent SDK.

8 levels · 64 lessons · 25 predict-and-verify challenges · 27 graded projects8 levels · 64 lessons · 25 predict-and-verify challenges · 27 graded projects · 20–30 hours

Curriculum

The 8 levels

64 lessons, 25 predict-and-verify challenges, 27 graded projects. Level 1 is open to everyone.

CapstoneAn End-to-End Automation

The agentic loop

gather · act · verify

the failure becomes the next lap's context

gather

read + search

act

edit + run

verify

check the result

FAIL

PASS

fix login

goal verified · loop stops

A failed verify is not a dead end. Its output becomes the context the next lap gathers, so the second attempt lands where the first missed, and the loop stops only when the check comes back green.

Level 1· Free

Usage: Chat, Cowork, and Code

Read a real session as a loop, route jobs to the right surface, install Claude Code and learn five moves most daily users miss, then ship a change from your own terminal.

4 lessons3 challenges2 projects~3h

1.1One sentence in, a verified fix out7m 1.2Chat, Cowork, or Code: route the job5m 1.3Route the job6m 1.4The loop, and how it recovers6m 1.5Read the transcript7m 1.6Install it, run one session, learn five moves8m 1.7What you don't know yet5m

Projects · build in your environment, submit for grading

1.P1Ship Your First Change with Claude Code45 min · rubric-graded 1.P2Onboard Onto an Unknown Codebase with Claude Code60 min · rubric-graded

Module quiz · 5 questions

Tailored to this module · graded inline with explanations

No account needed.

Outcomes

What you'll have built

Real artifacts from the projects, kept in your own setup after the course.

After Level 1

Your first shipped change, driven by Claude Code

A real edit in a real repository: explored, planned, implemented, and verified in your own terminal, with the session transcript to show for it.

After Level 5

A hardened .claude/ directory

Least-privilege permission rules, a custom slash command, a skill, and a PreToolUse hook that blocks secret-file writes, proven firing in a transcript.

After Level 6

A working MCP integration

An MCP server configured at the right scope and transport, with a transcript of Claude using its tools to complete a task.

After Level 8

A custom agent built on the Agent SDK

A small production-shaped agent with scoped tools, a turn budget, and schema-validated structured output.

Why this course

Built for people who already have Claude Code open

levels

graded projects

5 / 5

CCA-F domains covered

Do first, then read

Committing to an answer before the explanation is what makes the explanation stick, and it is how you already work with Claude Code: try it, watch what happens, adjust.

Pairs with Anthropic’s own primer

Claude Code in Action is a one-hour video introduction and we recommend it. This is the graded, hands-on program that goes deeper, with projects you keep in your own setup.

Aligned to the CCA-F exam

By Level 8 you have hands-on coverage of all five Claude Certified Architect (Foundations) domains, from Claude Code configuration to agentic architecture and context management.

CCA-F practice tests and flashcards

After the course

What you can do, and what you'll have used

You'll be able to

Drive the full Claude Code workflow: explore a codebase, plan a change in plan mode, implement it, and verify it before committing
Author CLAUDE.md hierarchies and path-scoped rules that steer Claude across a real multi-package repository
Configure least-privilege permissions, custom slash commands, skills, and lifecycle hooks in settings.json and .claude/
Connect Claude to external tools through MCP servers, choosing the right scope, transport, and auth model
Orchestrate subagents with isolated contexts, and run Claude headless in scripts and CI pipelines
Build a small working agent with the Claude Agent SDK, with scoped tools and structured output

You'll work with

Core

Claude Code CLICLAUDE.mdsettings.jsonplan modecheckpoints

Customization

Skillsslash commandshookspermission rules

Integration

MCP serverspluginsGitHub Actionsheadless mode

Building

Claude Agent SDK (Python & TypeScript)structured outputsubagents

You'll need: Comfortable in a terminal (cd, run commands, read output) · A Claude account; a Pro/Max subscription or API key to run Claude Code locally · Basic git for the later levels (branch, diff, commit)

FAQ

Questions, answered

The in-browser challenges run free right on the page. To do the verify steps and boss projects you need Claude Code running locally, which requires a Claude Pro/Max subscription or an Anthropic API key. Most learners already have one; that is the point of the course.

About this course. Claude Code is an agent that works in your terminal: it reads your codebase, edits files, runs commands, and verifies its own work. This course teaches it the way it deserves to be taught, hands-on. Each level pairs quick predict-then-verify challenges (commit your prediction, then prove it in your own terminal) with a boss project you complete in your real environment and submit for rubric grading. By Level 8 you are orchestrating subagents, wiring MCP servers, and building your own agent on the Agent SDK. The same levels double as hands-on preparation for the Claude Code, MCP, and agentic-architecture domains of Anthropic's CCA-F certification.

Reading

Guides that pair with the course

Learn Claude Code in 2026: The Complete 8-Level PathA structured path to learn Claude Code from first install to autonomous agents: 8 levels, what each one teaches, and how to practice every skill hands-on.Claude Code Tutorial for Beginners: Your First Real Session [2026]A beginner's Claude Code tutorial that goes beyond install: your first session, how the agentic loop works, the five commands that matter, and what to learn next.8 Projects That Teach You Claude Code, Level by Level [2026]Learn Claude Code by building: eight hands-on projects from first supervised session to an Agent SDK capstone, each mapped to the skill it forces you to learn.Is Claude Code Worth Learning in 2026? An Honest AssessmentWhether Claude Code is worth learning in 2026: who gets real returns, the learning curve by level, honest costs, and how the skill transfers across tools.CLAUDE.md & Context Management Best Practices (2026 Guide)How to write a CLAUDE.md that actually works: the memory hierarchy, imports, what belongs in persistent context vs skills, plus the context-window management principles the CCA-F exam tests.Claude Code Permissions & Settings Precedence (CCA-F Guide)How Claude Code's permission system and settings precedence actually work: allow/ask/deny rules, the settings hierarchy, permission modes, and the safety scenarios the CCA-F exam tests.Claude Code Skills & Slash Commands: When to Use Which (CCA-F)Claude Code skills vs slash commands vs CLAUDE.md vs MCP: what each mechanism is, how skills load on demand, how slash commands differ, and the selection questions the CCA-F Configuration & Workflows domain tests.Claude Code Hooks Explained: Lifecycle, Exit Codes & Exam Angles (CCA-F)Claude Code hooks from first principles: every lifecycle event, how exit codes control the agent, hooks vs CLAUDE.md instructions, and the enforcement scenarios the CCA-F exam tests in the Configuration & Workflows domain.MCP Tool Design Best Practices for Claude Agents (CCA-F)How to design tools Claude can actually use well: naming, descriptions, input schemas, error handling, and token-efficient results. The Tool Design & MCP Integration domain of the CCA-F exam, in practice.Claude Code Subagents & Orchestration Patterns (CCA-F Deep Dive)How Claude Code subagents actually work: separate context windows, agent definitions, delegation patterns, and the orchestration architectures tested in the CCA-F exam's largest domain (27%).Structured Output & Prompt Engineering for Claude (CCA-F)Reliable JSON from Claude, tool-use prompting, XML tags, prefilling, and the prompt-engineering techniques the CCA-F exam tests in its 20% Prompt Engineering & Structured Output domain.Model Context Protocol (MCP) Architecture Guide: From N×M to N+M [2026]How MCP eliminates the N×M integration tax in production agent stacks. Covers the 2026 stateless protocol revision, Streamable HTTP vs stdio, when to use MCP vs Anthropic Skills vs native tools, and the CCA-F + NCP-AAI exam angles.

Start with Level 1. It's free.

The other 7 levels, all 27 auto-graded projects, and every lab and practice test on Preporato come with Pro. $29.99/mo, cancel anytime.