Introductory week: understand the Claude model, the REST API and the main configuration parameters before advancing to more sophisticated techniques.
← Tracks
certification Claude Certified Architect — Foundations | Anthropic
CCA Foundation
Claude Certified Architect — Foundations | Anthropic · ~80 hours · 0 of 18 resources
Agentic (27%)
Claude Code (20%)
Prompt eng. (20%)
MCP & tools (18%)
Context (15%)
Building with the Claude API
Authentication, basic calls, streaming and roles (user/assistant/system).
Access →+ Complementary
Exercises
- Create an account in Anthropic Console and make the first API call via curl
- Explore the temperature, max_tokens and system prompt parameters
- Implement a basic API call script in Python or TypeScript
Building with the Claude API (advanced modules)
Advanced prompting techniques, XML tags, chain-of-thought and output control.
Access →Tool Use — official documentation
How to define and use tools to extract structured data and interact with external systems.
Access →+ Complementary
Structured output exercises (GitHub)
Practical examples and exercises for structured data extraction with Claude.
Access →Exercises
- Implement chain-of-thought prompting in 3 different problems
- Use tool use to extract structured data from free-form text
- Compare outputs with and without XML tags in the system prompt
Focus on prompt engineering techniques and structured data extraction — core competencies for the exam.
Claude Code in Action
Practical use of Claude Code in real projects — workflows and best practices.
Access →Claude Code — Official Documentation
Complete reference: hooks, slash commands, configuration, and IDE integration.
Access →+ Complementary
Exercises
- Install and configure Claude Code locally
- Use Claude Code to refactor an existing project by reviewing each change
- Create a custom skill and test it via slash command
Claude Code as an AI-assisted development tool — installation, daily use, and customization via skills.
Introduction to Subagents
Orchestration patterns with subagents — delegation, parallelism, and coordination.
Access →Claude Code — official documentation
Advanced configuration: pre/post tool call hooks, permissions, and workflow automation.
Access →+ Complementary
Agent SDK Cheat Sheet (Tutorials Dojo)
Quick reference for Agent SDK with code examples and common patterns.
Access →Exercises
- Implement a workflow with subagents using the Agent SDK
- Configure pre/post tool call hooks in Claude Code
- Create an agent that delegates subtasks to specialized subagents
Subagent orchestration and advanced Claude Code configuration — a topic with significant weight in the Agentic domain.
Introduction to Model Context Protocol
MCP fundamentals — architecture, server types and use cases.
Access →Tool Use — official documentation
Efficient tool design — schemas, descriptions and integration best practices.
Access →MCP Build Guide (official)
Practical guide to building an MCP server from scratch with support for resources and tools.
Access →+ Complementary
Exercises
- Build a simple MCP server with 2-3 custom tools
- Connect the MCP server to Claude Code and test the tools
- Implement an MCP resource server that exposes data from an external API
Model Context Protocol: how to expose tools and context to Claude in a structured and reusable way.
MCP Advanced Topics
Advanced MCP topics — sampling, authentication, and composition patterns.
Access →Agents & Tools — official documentation
Agentic architecture patterns — orchestration, memory, and reliable systems design.
Access →+ Complementary
Exercises
- Design the architecture of a multi-step agent for a real-world problem
- Implement memory and context management in an agent
- Create an orchestrator that coordinates specialized agents in parallel
The Agentic domain is worth 27% of the exam — focus on architecture patterns, orchestration, and reliable systems design.
Prompt Engineering — official documentation
Context management: caching, context windows and token optimization in long conversations.
Access →Develop Tests — official documentation
How to build eval suites, quality metrics and automated testing pipelines.
Access →+ Complementary
Practice Questions (claudecertifications.com)
Question bank in exam style to test knowledge by domain.
Access →Exam Guide (claudecertifications.com)
Official exam guide with learning objectives and weights by domain.
Access →Community Exam Guide (GitHub)
Collaborative community guide with tips from those who have taken the exam.
Access →Exercises
- Implement an automated eval suite for a critical prompt
- Measure and optimize context window usage in long conversations
- Solve 25 practice questions and review the errors
Context management and quality assessment — how to ensure that systems with Claude work consistently and measurably.
Anthropic Academy — official practice exam
Review of all 5 domains via practice exams and exercises on Anthropic's official platform.
Access →+ Complementary
Complete Study Guide (GitHub)
Comprehensive study guide with questions and code examples by domain.
Access →Agent SDK Cheat Sheet (Tutorials Dojo)
Quick reference for Agent SDK to review orchestration concepts.
Access →CertificationPractice — full-length practice exam
Full-length practice exam with timer and detailed feedback per question.
Access →Exercises
- Complete the full practice exam on CertificationPractice and identify weak points
- Review the 2 domains with the lowest scores in practice exams
- Retake the 25 practice questions aiming for 100% accuracy
Consolidation week. Goal: score >900/1000 on the exam. Focus on the gaps identified in previous practice exams.