CCAO-F cheat sheet

Exam at a glance

Domain weightings

How much of the exam each domain covers. Spend your study time in proportion — the heavier the domain, the more questions you'll see.

Key terms

Messages API

The Messages API is Anthropic's primary interface for conversing with Claude, where you send an ordered list of user and assistant messages and receive Claude's reply. It supports system prompts, streaming, tool use, and parameters such as max_tokens and temperature.

System Prompt

A system prompt is the instruction, passed separately from the conversation, that sets Claude's role, behaviour, and constraints for the whole exchange. Giving Claude a clear role in the system prompt is a core prompt-engineering technique.

Context Window

The context window is the maximum amount of text, measured in tokens, that Claude can consider at once — including the system prompt, the conversation, and the response. Staying within it, and choosing a model with a large enough window, is essential for long inputs.

Token

A token is the basic unit of text that Claude processes, roughly a few characters or part of a word. Both input and output are measured in tokens, which drives cost, latency, and whether a request fits in the context window.

Prompt Engineering

Prompt engineering is the practice of designing inputs — instructions, examples, and structure — to reliably get the output you want from Claude. Techniques include being clear and direct, using examples, chain-of-thought, XML tags, and prompt chaining.

Chain of Thought

Chain of thought is a prompting technique that asks Claude to reason step by step before giving a final answer, which improves accuracy on complex tasks. It can be elicited by instruction or by giving Claude space to think in the response.

Multishot Prompting

Multishot prompting is the technique of including several examples of the desired input-output behaviour in the prompt so Claude generalises the pattern. It is one of the most effective ways to steer format and quality.

Tool Use

Tool use (function calling) is the capability that lets Claude call external functions you define with a name, description, and input schema. Claude decides when to request a tool, you run it and return the result, and Claude incorporates the result into its answer.

Model Context Protocol

The Model Context Protocol (MCP) is an open protocol that standardises how applications connect Claude to external tools and data sources. MCP servers expose tools and context, and MCP clients let Claude use them without custom integration for each source.

Agent

An agent is a system in which Claude directs its own control flow — deciding which tools to call and when — to accomplish a goal, as opposed to a workflow that follows predefined steps. Agents suit open-ended tasks but need bounds, tools, and feedback loops to stay reliable.

Prompt Caching

Prompt caching is a feature that stores a reusable prefix of a prompt so repeated requests reuse it, reducing latency and cost. It is valuable when a large, stable context (such as a document or system prompt) is sent on many requests.

Hallucination

A hallucination is a confident but incorrect or fabricated output from a language model. Grounding Claude in provided sources, allowing it to say it does not know, and asking for citations are practical techniques to reduce hallucinations.

Prompt Injection

Prompt injection is an attack in which untrusted input tries to override the developer's instructions to Claude, for example by hiding commands inside a document or webpage. Input/output validation, guardrails, and keeping a human in the loop help mitigate it.