This course is based on the official Claude Certified Architect (Foundations) exam guide. The exam itself is only available to Claude partners. But that does not matter. The knowledge is what counts, and the knowledge is what this course gives you.
01 The Exam You Cannot Take (And Why That Is Fine)
The Claude Certified Architect (Foundations) exam tests your ability to build production-grade applications using four core technologies:
| Technology | What It Does |
|---|---|
| Claude Code | Anthropic's official CLI for AI-assisted development |
| Claude Agent SDK | Framework for building agentic applications |
| Claude API | The Messages API for programmatic Claude access |
| Model Context Protocol (MCP) | Open protocol for connecting AI to external tools and data |
To take the exam, you need to be a Claude partner. To build production-grade applications, you just need to understand the material.
Every skill in this course is monetisable. You are not learning theory. You are learning how to build systems that companies pay for.
02 What the Exam Tests
The exam is scenario-based multiple choice. One correct answer, three plausible distractors. Passing score is 720 out of 1000. It consistently rewards:
- Deterministic solutions over probabilistic ones when stakes are high
- Proportionate fixes (simple first, complex only when needed)
- Root cause tracing (find the actual problem, not the symptom)
The six production scenarios the exam uses:
| Scenario | Skills Tested |
|---|---|
| Customer Support Resolution Agent | Agent SDK + MCP + escalation patterns |
| Code Generation with Claude Code | CLAUDE.md + plan mode + slash commands |
| Multi-Agent Research System | Coordinator-subagent orchestration |
| Developer Productivity Tools | Built-in tools + MCP servers |
| Claude Code for CI/CD | Non-interactive pipelines + structured output |
| Structured Data Extraction | JSON schemas + tool_use + validation loops |
03 The Five Domains
This course is structured around the five exam domains, weighted by importance:
| Domain | Weight | What It Covers |
|---|---|---|
| 1. Agentic Architecture & Orchestration | 27% | Agentic loops, multi-agent systems, hooks, session management |
| 2. Tool Design & MCP Integration | 18% | Tool descriptions, error handling, MCP servers, built-in tools |
| 3. Claude Code Configuration & Workflows | 20% | CLAUDE.md hierarchy, skills, plan mode, CI/CD |
| 4. Prompt Engineering & Structured Output | 20% | Explicit criteria, few-shot, tool_use, batch processing |
| 5. Context Management & Reliability | 15% | Context preservation, escalation, error propagation |
Domain 1 is the single most important domain. If you nail agentic architecture, you have a strong foundation for everything else.
04 How to Use This Course
Each domain chapter follows the same structure:
- Concept explanation with concrete production examples
- Exam traps — the specific anti-patterns and misconceptions tested
- Practice scenarios — real situations to test your understanding
- What to build — hands-on exercises that cover the domain
The final chapter gives you build exercises that tie everything together.
The single best way to learn this material is to build while you read. Do not just consume. Open Claude Code, open the Agent SDK docs, and build alongside each chapter.
05 Prerequisites
You should be comfortable with:
- Basic programming (Python or TypeScript)
- Command-line usage
- API concepts (REST, JSON, HTTP)
- A general understanding of how LLMs work
You do not need prior experience with Claude specifically. This course starts from the fundamentals of each technology.
06 Recommended Learning Resources from Anthropic
Before or alongside this course, work through these official materials:
- Building with the Claude API — Core API mechanics
- Introduction to Model Context Protocol — MCP fundamentals
- Claude Code in Action — Practical CLI usage
- Claude 101 — General Claude capabilities
Now let us get into the first and most important domain: Agentic Architecture.