Objective 2.2 · Domain 2 · 13% of the exam

Design system prompts, templates, and guardrails

System prompts, reusable templates and prompt-level guardrails, which is to say establishing role, scope, output format and refusal behaviour in the place that persists across a conversation. The distinction the exam draws is between instructions belonging in the system prompt, which are stable, behavioural and security-relevant, and instructions belonging in the user turn, which are task-specific and variable. Putting per-request data in a system prompt is a recurring wrong answer.

5 questions · answers and explanations shown as you go · free, no sign-up

The rest of domain 2: Models, Prompting & Context Engineering

This domain is 13% of a scored form, about 8 of the 63 questions, spread across 5 objectives. Drill the whole domain or pick another objective below.

2.1Select appropriate Claude models based on trade-offs(6)
2.3Apply prompt engineering techniques (zero-shot, few-shot, chain-of-thought)(5)
2.4Optimize context windows and manage token usage(5)
2.5Implement prompt reuse strategies (caching, modular prompts, Skills)(5)

One objective is a narrow slice. A full-length timed mock is what tells you whether the whole thing holds together.

Sit a timed mock →