Study guide

5.1 Prompt Engineering

Claude Certified Architect - Foundations Study Guide Exam

Track your progress as you read

Finished sections get a check mark.

1 Chapter 1: Exam and Ecosystem Overview

1.1 Certification Context 1.2 Study Map

2 Chapter 2: Agentic Architecture and Orchestration

2.1 Workflows vs Agents 2.2 Agentic Loops and the Claude Agent SDK 2.3 Multi-Agent Orchestration and Session Management

3 Chapter 3: Claude Code Configuration and Workflows

3.1 Claude Code Workflows

4 Chapter 4: Tool Design and MCP Integration

4.1 Tool Design and MCP Integration

5 Chapter 5: Prompt Engineering and Structured Output

5.1 Prompt Engineering 5.2 Structured Output and Evaluation

6 Chapter 6: Context Management and Reliability

6.1 Context, Caching, and Compaction 6.2 Model Selection and Reliability

5.1 Prompt Engineering

Key Takeaways Explain This Topic AI

A production prompt is a contract stating role, task, context, constraints, output shape, and success criteria — not a polite request.

Claude is tuned to respect XML-style delimiters (e.g. <document>...</document>) and weights explicit rules placed near the end of the prompt more strongly.

Few-shot examples teach format and edge cases; three or four diverse examples including a tricky case beat ten near-duplicates, and examples never replace code validation.

Prefer being prescriptive and explicit over clever: state what to do and what not to do, give the model an escape hatch (say 'I don't know'), and put dynamic context in the right place.

Advanced techniques (chain-of-thought / thinking at a higher effort, prompt chaining, role prompting) are tools to deploy when the task complexity justifies them, not defaults.

Last updated: October 2026

A Prompt Is a Contract

The CCA-F treats a prompt as a contract, not a polite request, and Prompt Engineering & Structured Output is a 20% domain. A production-grade prompt names six things:

  • Element What it specifies Example
  • Role Who Claude is acting as "You are a senior compliance analyst."
  • Task The single outcome "Summarize the loan application's risk factors."
  • Context The data to work from The application text in <application>...</application> tags
  • Constraints Length, tone, forbidden actions "Under 150 words. Do not infer missing income."
  • Output shape An explicit schema or format A JSON object with named fields
  • Success criteria How a grader would judge it "Every risk cited must point to a line in the application."

Naming all six removes ambiguity. The most common weak-prompt failure the exam describes is a vague instruction ("summarize this nicely") that leaves the output shape and the failure behavior undefined, so downstream code cannot rely on it.

Structure Claude Responds To

Two formatting habits measurably improve adherence and are worth knowing for the exam:

XML-style delimiters. Claude is tuned to respect tags such as <document>...</document> , <example>...</example> , and <instructions>...</instructions> . Wrapping each part of the prompt in a labeled tag prevents the model from confusing the data with the instructions — important when the data itself contains instruction-like text (a defense against prompt injection from retrieved content).

Rules near the end. A short, explicit list of rules placed near the end of the prompt benefits from recency and is followed more reliably than the same rules buried in the middle.

Be prescriptive and explicit rather than clever. State both what to do and what not to do. Give the model an escape hatch : tell it to say "I don't know" or return a sentinel value when the context lacks the answer, rather than forcing a fabricated response. And place dynamic context deliberately — durable rules go in the system prompt, the immediate task goes at the bottom, and anything that changes per request goes late so it does not invalidate prompt caching of the stable prefix (Chapter 6).

A further habit the exam rewards is matching prompt specificity to the task: positive instructions that show the desired behavior outperform long lists of prohibitions, and concrete success criteria ("each risk must cite a line number") beat vague adjectives ("be thorough"). When a recent Claude model follows instructions very literally, an over-prescriptive prompt can backfire by suppressing useful judgment — so state the goal and the hard constraints, then let the model fill in the rest rather than scripting every step.

Few-Shot Examples: Powerful but Not Validation

Examples are the strongest lever for teaching format and edge cases . The exam expects a few principles:

Diversity beats volume. Three or four diverse examples that include a tricky edge case outperform ten near-duplicates that all look the same. Duplicates teach the model nothing new and waste context budget.

Show the edge case. If the real input distribution includes empty fields, ambiguous requests, or adversarial text, an example should demonstrate the desired handling for at least one of them.

Examples never replace validation. A model can produce convincing prose that violates the contract. If downstream code parses the output, you still validate the parsed object in code (Chapter 5.2). Treating examples as a guarantee is a trap the exam punishes.

Wrap each example in tags ( <example> ) and, where it helps, label the input and the expected output so the pattern is unambiguous.

Advanced Techniques: Deploy When Justified

The exam expects familiarity with advanced techniques and, just as importantly, the judgment to apply them only when the task warrants it.

Technique What it does Use when

Chain-of-thought / thinking (raise effort) Let the model reason step by step before answering Multi-step reasoning, math, complex analysis

Role / persona prompting Set a role to shape tone and domain framing The task benefits from a specific voice or expertise

Prompt chaining Split a complex prompt into focused sequential calls One mega-prompt is unreliable; stages are clearer separately

Multi-pass review Generate, then critique/revise in a second pass Quality matters and a single pass under-delivers

Prefilling / format steering Guide the start of the response toward a shape Steering output format on models that support it

Modern Claude models expose extended (adaptive) thinking , where the model decides how much to reason; you control depth with an effort setting rather than a fixed token budget. The architect's discipline is the same as everywhere on this exam: reach for the simplest technique that meets the quality bar. A clear single prompt with good structure beats an elaborate chain when the task is simple, and the exam will offer an over-engineered prompting answer as a distractor.

Prompt Engineering Decision Table

  • Scenario clue Best move
  • Output must be machine-parsed Define the schema first; specify the exact output shape
  • Data contains instruction-like text Wrap data in XML tags; keep instructions separate
  • Model fabricates when unsure Add an explicit "say I don't know" escape hatch
  • Format is inconsistent Add three or four diverse few-shot examples
  • Complex multi-step reasoning Raise effort or use chain-of-thought
  • One huge prompt is unreliable Split into a prompt chain of focused steps

Match the clue to the lever, and prefer the least elaborate prompt that satisfies the success criteria.