Exam overview
The Claude Certified Architect — Foundations (CCA) certification validates that practitioners can make informed architectural decisions when implementing real-world solutions with Claude. Launched March 12, 2026, it is Anthropic's first official technical credential.
Exam format
Format
Multiple choice, 1 correct answer
Passing score
720 / 1,000 (scaled)
Scenarios
4 of 6 selected randomly
Access
Claude Partner Network
Domain weights
Domain 1Agentic Architecture & Orchestration27%
Domain 3Claude Code Configuration & Workflows20%
Domain 4Prompt Engineering & Structured Output20%
Domain 2Tool Design & MCP Integration18%
Domain 5Context Management & Reliability15%
The 6 exam scenarios
Four scenarios are randomly selected at exam time. Every question is anchored to one of the selected scenarios. You are the architect of a specific production system — not answering in the abstract.
Scenario 1Customer Support Resolution Agent
Building a support agent using the Claude Agent SDK with MCP tools (get_customer, lookup_order, process_refund, escalate_to_human). Target: 80%+ first-contact resolution.
D1D2D5
Scenario 2Code Generation with Claude Code
Using Claude Code for code generation, refactoring, debugging, and documentation with CLAUDE.md configurations, custom slash commands, and plan mode.
D3D5
Scenario 3Multi-Agent Research System
A coordinator agent delegating to specialist subagents: web search, document analysis, synthesis, and report generation. Produces comprehensive cited reports.
D1D2D5
Scenario 4Developer Productivity Tools
An agent helping engineers explore codebases, understand legacy systems, generate boilerplate, and automate repetitive tasks using built-in tools and MCP servers.
D1D2D3
Scenario 5Claude Code for CI/CD
Integrating Claude Code into CI/CD pipelines for automated code review, test generation, and PR feedback. Designing prompts that minimize false positives.
D3D4
Scenario 6Structured Data Extraction
Extracting structured information from unstructured documents, validating output with JSON schemas, and handling edge cases gracefully with retry loops.
D4D5
Target candidate
The ideal candidate has 6+ months of hands-on production experience with:
- Building agentic applications using the Claude Agent SDK (multi-agent orchestration, tool integration, hooks)
- Configuring Claude Code for team workflows using CLAUDE.md files, Agent Skills, and MCP server integrations
- Designing Model Context Protocol (MCP) tool and resource interfaces for backend system integration
- Engineering prompts that produce reliable structured output using JSON schemas, few-shot examples, and validation patterns
- Managing context windows effectively across long documents, multi-turn conversations, and multi-agent handoffs
- Integrating Claude into CI/CD pipelines for automated code review, test generation, and pull request feedback
6-week study plan
Week 1–2
Claude API & core concepts
Complete Anthropic Academy courses on Skilljar. Master the API fundamentals: messages, tools, stop_reason, streaming. Build a simple agentic loop from scratch.
Week 3
Agent SDK & multi-agent orchestration
Study the Agent SDK documentation. Build a coordinator-subagent system. Practice with the Task tool, hooks, and context passing patterns. Focus on Domain 1.
Week 4
MCP & tool design
Learn MCP tool interface design: descriptions, error structure (isError, errorCategory, isRetryable), tool scoping. Configure .mcp.json. Focus on Domain 2.
Week 5
Claude Code & prompt engineering
Configure CLAUDE.md hierarchies, set up custom slash commands and hooks. Practice structured output with JSON schemas and validation retry loops. Domains 3 & 4.
Week 6
Context management + mock exam
Study context strategies and escalation patterns. Complete the full 100-question practice set on this site. Take the official 60-question practice exam before booking.
Official resources