Study guides / CCAR-P
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 - Integration
-
Audit and prune a 40-tool agent configuration
From 1.1 - Auditing Tool and Agent Configuration for Capability Bloat
-
Threat-model the credential flow of a support agent
From 1.2 - Authentication and Authorisation Gaps in Claude Integrations
-
Sweep tier and effort and write a defensible configuration record
From 1.3 - Accuracy-Latency Trade-offs and Defending the Configuration
-
Instrument a multi-agent flow with correlated, privacy-aware telemetry
From 1.4 - Observability Challenges at Scale and Choosing a Monitoring Strategy
-
Chunk, enrich, evaluate and cite a small knowledge base
From 1.5 - Designing a RAG Pipeline: Chunking and Indexing
-
Run a retrieval bake-off across query patterns
From 1.6 - Retrieval Strategies Matched to Data Shape and Query Pattern
-
Choose and document integration mechanisms for three capabilities
From 1.7 - Choosing an Integration Mechanism: MCP, API/CLI, or Agent-to-Agent
-
Convert a monolithic tool set to deferred loading and measure the trade-off
From 1.8 - Progressive Discovery vs a Monolithic Context Strategy
Domain 2 - Solution Design & Architecture
-
Write a use-case fit brief and go/no-go plan
From 2.1 - Translating a Business Problem into a Claude-Based Solution
-
Build a validated pipeline with a closed feedback loop
From 2.2 - End-to-End Architecture: Input, Processing, Output and Feedback Loops
-
Prototype one task three ways and write the decision record
From 2.3 - Workflow, Agentic, or Augmented LLM: Choosing the Pattern
-
Test whether multi-agent earns its cost
From 2.4 - Multi-Agent Systems and Orchestration Strategy
-
Decompose a portfolio review and prove it beats one big pass
From 2.5 - Decomposition Techniques for Complex Problems
-
Trace value to design and measure the trade-offs
From 2.6 - Aligning a Design to Business-Value Pillars and SLAs
Domain 3 - Evaluation, Testing & Optimisation
-
Write a metric specification and grade it: code, model and latency metrics for one use case
From 3.1 - Defining Evaluation Metrics: Accuracy, Latency, Cost, Safety and Security
-
Build a versioned, sliced evaluation dataset with a held-out set and a calibrated judge
From 3.2 - Evaluation Datasets and Mixed-Methodology Test Frameworks
-
Plan and simulate a safe A/B comparison: sticky bucketing, a paired offline comparison and a rollout rule
From 3.3 - A/B Testing and Iterative Improvement
-
Build a failure-triage kit: evidence capture, layer ablations and a failure profile
From 3.4 - Diagnosing Failures: Prompt Failure, Hallucination and Model Mismatch
-
Profile, optimise and defend: caching, batching and an evaluated tier or effort change
From 3.5 - Optimising Token Usage, Latency and Cost-Performance
-
Build a small quality-and-cost monitoring loop: log, roll up, replay, alert and sample
From 3.6 - Production Monitoring with Logging and Observability Tooling
Domain 4 - Governance, Safety & Risk Management
-
Design and test a layered guardrail set for an email-triage agent
From 4.1 - Implementing Guardrails and Safety Controls
-
Build a risk register for an LLM workflow and measure one risk
From 4.2 - Risks, Limitations and Failure Modes of LLM Systems
-
Design and simulate a review policy with routing, sampling and reviewer-vigilance metrics
From 4.3 - Human-in-the-Loop Validation Strategies
-
Produce a compliance-oriented architecture pack: data-flow map, residency decision record, feature audit and a minimisation gateway
From 4.4 - Compliance in the Architecture: GDPR, HIPAA and FedRAMP
-
Build a counterfactual bias test and the transparency artifacts around it
From 4.5 - Ethical AI: Bias, Fairness and Transparency
Domain 5 - Stakeholder Communication & Lifecycle Management
-
Design a discovery session and turn the answers into architecture implications
From 5.1 - Structured Discovery and Requirements Gathering
-
One decision, three audiences, one decision record
From 5.2 - Communicating Architectural Decisions and Trade-offs
-
Write an expectations and SLO charter with a feedback loop
From 5.3 - Stakeholder Feedback Loops and Expectation and SLA Alignment
-
Draft an architecture document skeleton, a dynamic diagram, a runbook and a handoff checklist
From 5.4 - Documenting an Architecture and Giving Implementation Guidance
-
Plan and rehearse a model migration and write the lifecycle ownership table
From 5.5 - Supporting the Lifecycle: Discovery, Design, Handoff, Monitoring, Iteration
Domain 6 - Claude Models, Prompting & Context Engineering
-
Write a model-selection decision record backed by a cost-per-completed-task eval
From 6.1 - Selecting a Claude Model on Cost, Latency and Capability Trade-offs
-
Harden a system prompt and prove it with an eval
From 6.2 - System Prompts, Templates and Guardrails
-
Run an ablation: zero-shot, few-shot and thinking on an extraction task
From 6.3 - Prompting Techniques: Zero-Shot, Few-Shot and Chain-of-Thought
-
Instrument and shrink the context of a tool-heavy agent
From 6.4 - Optimising the Context Window and Managing Token Usage
-
Cache it, modularise it, package it: measure savings and write the governance record
From 6.5 - Prompt Reuse: Caching, Modular Prompts and Skills
Domain 7 - Developer Productivity & Operational Enablement
-
Design, stage and prove a team Claude Code configuration
From 7.1 - Configuring Claude Code and Environments for a Team
-
Build and evaluate an AI-assisted delivery workflow
From 7.2 - AI-Assisted Developer Workflows
-
Fault-injection lab and escalation runbook
From 7.3 - Debugging and Operational Issue Resolution