Study guides / CCAR-F
Build exercises
A hands-on exercise for every lesson - one build task, with hints, to prove you can actually apply the concept, not just recognise it.
Domain 1 - Agentic Architecture & Orchestration
-
Build a Multi-Tool Agent Loop
From 1.1 - Agentic Loops
-
Build a Hub-and-Spoke Research Coordinator
From 1.2 - Multi-Agent Orchestration
-
Implement Context Passing with Structured Metadata
From 1.3 - Subagent Invocation and Context Passing
-
Build a Prerequisite Gate for Financial Operations
From 1.4 - Workflow Enforcement and Handoff
-
Implement Agent SDK Hooks for Normalisation and Policy Enforcement
From 1.5 - Agent SDK Hooks
-
Build a Multi-Pass Code Review Pipeline
From 1.6 - Task Decomposition Strategies
-
Implement Session Management Strategies
From 1.7 - Session State and Resumption
Domain 2 - Tool Design & MCP Integration
-
Design Tool Descriptions That Eliminate Misrouting
From 2.1 - Tool Interface Design
-
Build Structured Error Responses for All Four Categories
From 2.2 - Structured Error Responses
-
Configure Tool Distribution Across a Multi-Agent System
From 2.3 - Tool Distribution & Tool Choice
-
Configure MCP Servers with Scoping and Environment Variables
From 2.4 - MCP Server Integration
-
Trace and Refactor a Deprecated Function Using Built-in Tools
From 2.5 - Built-in Tools
Domain 3 - Claude Code Configuration & Workflows
-
Build a Multi-Level CLAUDE.md Configuration
From 3.1 - CLAUDE.md Hierarchy, Scoping, and Modular Organisation
-
Create Custom Commands and Skills
From 3.2 - Custom Slash Commands and Skills
-
Configure Path-Specific Rules with Glob Patterns
From 3.3 - Path-Specific Rules for Conditional Convention Loading
-
Practice Plan Mode vs Direct Execution Decision-Making
From 3.4 - Plan Mode vs Direct Execution
-
Practice Iterative Refinement Techniques
From 3.5 - Iterative Refinement Techniques
-
Set Up a CI/CD Pipeline with Claude Code
From 3.6 - CI/CD Integration
Domain 4 - Prompt Engineering & Structured Output
-
Build an Explicit Criteria Code Review Prompt
From 4.1 - System Prompts with Explicit Criteria
-
Build a Few-Shot Enhanced Extraction Prompt
From 4.2 - Few-Shot Prompting
-
Build a Structured Extraction Tool with JSON Schema
From 4.3 - Structured Output with Tool Use
-
Build a Validation-Retry Loop for Document Extraction
From 4.4 - Validation, Retry, and Feedback Loops
-
Design a Batch Processing Strategy
From 4.5 - Batch Processing Strategies
-
Build a Multi-Pass Code Review System
From 4.6 - Multi-Instance and Multi-Pass Review
Domain 5 - Context Management & Reliability
-
Build a Persistent Case Facts Context Manager
From 5.1 - Context Window Management
-
Build an Escalation Decision Engine
From 5.2 - Escalation & Ambiguity Resolution
-
Build a Structured Error Propagation System
From 5.3 - Error Propagation in Multi-Agent Systems
-
Build a Context-Resilient Codebase Explorer
From 5.4 - Codebase Exploration & Context Degradation
-
Build a Confidence-Calibrated Review Router
From 5.5 - Human Review & Confidence Calibration
-
Build a Provenance-Preserving Synthesis Pipeline
From 5.6 - Information Provenance & Multi-Source Synthesis