Glossary
Plain-English definitions, no prior knowledge assumed. Skim it, or search for a term you hit elsewhere.
Agent
An AI set up to take actions in a loop — it reads a goal, calls tools, sees the results, and repeats until done. Claude Code is an agent for your codebase.
Anthropic
The company that makes Claude.
API (Application Programming Interface)
A way for your software to talk to Claude programmatically, instead of typing in a chat window. See Your First API Call.
Artifact
A live, runnable output (an app, chart, or document) that Claude renders in a side panel on claude.ai, separate from the chat.
Chain-of-thought (CoT)
Asking the model to reason step by step before answering, which improves accuracy on hard problems.
Claude
The family of AI models (and apps) made by Anthropic. "Claude" can mean the chat app, the models, or the brand depending on context.
Claude Code
Anthropic's agentic command-line/IDE tool that can read and edit your codebase, run commands, and use tools. See What Claude Code Is.
CLAUDE.md
A plain-text file where you give Claude Code persistent instructions about your project — its highest-leverage customisation.
Compaction
Automatically condensing older conversation turns into a short summary when the context window fills, so a long session can keep going without losing the thread. See Context Management.
Context engineering
Deliberately deciding what goes into the model's context window — instructions, examples, retrieved data — to get the best result. The discipline behind prompting, RAG, and memory.
Context window
The maximum amount of text (measured in tokens) a model can consider at once — your prompt plus its reply. Exceed it and the oldest content drops off.
Effort
A control on newer Claude models for roughly how much the model thinks before answering. Match the tier to the task instead of maxing it out — you pay for thinking in latency and cost. See Extended Thinking & Effort.
Embedding
A list of numbers representing the meaning of a piece of text, so software can find things that are similar in meaning (not just matching keywords). Powers semantic search and RAG.
Eval (evaluation)
A repeatable test of AI output quality, so you can tell whether a change made things better or worse.
Few-shot
Including a few worked examples in your prompt to show the model the pattern you want.
Fine-tuning
Further-training a model on your own examples to specialise it. Usually a last resort — try prompting and RAG first.
Grounding
Instructing the model to answer only from the context you provide — and to say "I don't know" when the answer isn't there. The main prompt-level defense against hallucination. See the Pattern Library.
Guardrails
Rules and limits that keep an AI's behaviour safe and on-task — e.g. permission rules, "run tests before done", or human approval for risky actions. See Securing Agents.
Hallucination
When a model states something false but confident. The core skill is verification. See Hallucinations.
Hook
In Claude Code, a command that runs automatically at a lifecycle event (e.g. after every file edit) to enforce formatting, gates, or safety.
Inference
The act of running a model to get an output (as opposed to training it).
Latency
How long you wait for a response. Bigger models are usually slower; streaming improves perceived latency by showing output as it's generated.
LLM (Large Language Model)
The kind of AI behind Claude: trained to predict the next chunk of text, which turns out to be enough to write, code, summarise, and reason. See What Is an LLM?.
MCP (Model Context Protocol)
An open standard for connecting AI to external tools and data (databases, GitHub, browsers). An "MCP server" exposes a tool; Claude can call it. See MCP.
Multimodal
A model that handles more than text — e.g. images and PDFs as input.
Output style
In Claude Code, a preset that changes how Claude communicates (e.g. concise, explanatory, learning) without changing what it can do. See Output Styles.
Permission mode
In Claude Code, how much it can do without asking — from prompting on every action to auto-accepting edits. The main speed-vs-safety dial. See Permissions.
Plan mode
A read-only Claude Code mode that researches and proposes a plan before touching any files — you approve before it acts. See Plan Mode.
Prompt
What you send the model: your instructions, question, and any context.
Prompt caching
Reusing the unchanged front part of a prompt across calls to cut cost and latency. See Prompt Caching.
Prompt injection
An attack where malicious instructions hidden in content the AI reads (a web page, a document) try to hijack it. See Prompt Injection.
RAG (Retrieval-Augmented Generation)
Fetching relevant snippets from your data and adding them to the prompt so the model answers about them. See RAG.
Reasoning / extended thinking
A mode where the model spends extra compute "thinking" before answering, for harder tasks. On newer models you tune it with effort.
Skill
A packaged capability (instructions + optional scripts) that Claude loads on demand when relevant. See Skills.
Slash command
A /-prefixed shortcut in Claude Code — built-in (like /clear) or your own reusable prompt — that triggers a defined action. See Slash Commands.
Structured output
Getting results in a strict, machine-readable shape (e.g. JSON matching a schema) instead of free text — so other software can rely on them. See Structured Output.
Subagent
A separate Claude instance with its own context and tools, used to delegate part of a task (e.g. research) so the main session stays focused.
System prompt
The high-level instructions that set the model's role and rules for a whole conversation — your most powerful steering lever.
Temperature
A setting (on some models) controlling randomness: low = focused and repeatable, high = varied and creative.
Token
The unit models read and count in — roughly ¾ of a word in English. Pricing and context limits are measured in tokens. See Tokens & Context.
Tool use / function calling
Letting the model call functions you define (search, calculator, your API) and use the results. See Tool Use.
Missing a term? Add it — glossary entries are a perfect first contribution.