Skip to content

Glossary

Plain-language definitions of the terms used across this guide. If a page throws a word at you that you don't recognize, it's probably here.

API

A way for one program to ask another program for something. When your app "calls an API," it sends a request over the internet and gets data back. Example: your weather app calls a weather service's API to get today's forecast.

Auth

Short for authentication and authorization: proving who a user is (login) and deciding what they're allowed to do. Getting this wrong leaks people's accounts, so you use a service instead of building it yourself. Example: Clerk, Supabase Auth.

CDN

Content delivery network: a fleet of servers spread around the world that keep copies of your files (images, scripts, pages) close to each user, so they load fast everywhere instead of from one distant origin. Most hosts and upload services include one. Example: Cloudflare serving your site from the city nearest each visitor.

CI/CD

Continuous Integration / Continuous Deployment: robots that automatically run your tests and ship your code every time you push to GitHub, so "it works on my machine" stops being a sentence you say. Example: GitHub Actions.

CLI

Command-line interface: a text-based way to control programs by typing commands instead of clicking buttons. The black terminal window. Example: typing git push to upload your code.

Database

The organized place your app stores data so it survives after you close the browser tab — users, posts, orders, anything. Example: Postgres (often via Supabase).

Deployment

Putting your app on the public internet so other people can actually use it at a real URL, instead of it only running on your laptop. Example: pushing to Cloudflare and getting myapp.workers.dev.

Environment variable (.env)

A setting your code reads at runtime — like an API key or database password — kept in a separate .env file instead of hard-coded in your source. You keep this file out of version control so you don't leak secrets. Example: OPENAI_API_KEY=sk-... in a .env file.

Error monitoring

A service that catches crashes and bugs that happen for real users in production and tells you about them — with the exact line of code — so you don't find out from an angry tweet. Example: Sentry.

Formatter

A tool that automatically rewrites your code into a consistent style (spacing, quotes, line breaks) every time you save. You stop arguing with yourself about formatting and never ask the AI to "make it neat" again. Example: Biome, Prettier.

Framework

A pre-built skeleton for your app that handles the common, boring parts (routing, rendering, structure) so you only write the parts unique to your idea. Example: Next.js for web, Expo for mobile.

Git / GitHub

Git is the tool that tracks every change to your code so you can undo mistakes and see history; GitHub is the website where you store that code online and collaborate. Git is the engine, GitHub is the garage. Example: git commit saves a snapshot; pushing it to GitHub backs it up.

Linter

A tool that reads your code and flags likely bugs and bad patterns before you run it — like spellcheck for logic. Different from a formatter: a formatter fixes how code looks, a linter warns about what it does. Example: Biome, ESLint.

LLM

Large language model: the AI behind assistants like Claude — you send it text (a "prompt") and it sends text back. An "AI app" is usually a normal app that makes an API call to one. Example: Claude turning your chat messages into a reply.

ORM

Object-Relational Mapper: a layer that lets you talk to your database using normal code instead of writing raw SQL strings by hand, with autocomplete and type safety. Example: Drizzle.

Package manager

A tool that downloads and tracks the third-party code libraries your project depends on, and keeps everyone on the same versions. Without it you'd copy-paste libraries by hand and break things. Example: pnpm.

RAG

Retrieval-Augmented Generation: a technique where you fetch relevant documents first and hand them to an LLM as context, so it answers from your data instead of making things up. Example: a chatbot that answers questions about your company docs by looking them up before replying.

SDK

Software development kit: a ready-made library that wraps a service's API so you call it with a few lines of typed code instead of hand-building web requests. Example: the Anthropic SDK for calling Claude, or the AI SDK for streaming chat UIs.

Serverless

A hosting model where the platform runs your code on demand and scales it for you — you never provision or babysit a server, and you pay only for what actually runs. Example: Cloudflare Workers running your API without you managing any machine.

Type checker / TypeScript

TypeScript is JavaScript with labels on your data ("this is a number, this is a user"), and the type checker is the tool that catches when those labels don't match — finding whole categories of bugs before you run the code. Example: passing a string where a number was expected gets flagged as you type.

Vector database

A database built to store the "meaning" of text as long lists of numbers (embeddings) and find the most similar items fast — the backbone of search and RAG in AI apps. Example: Supabase pgvector.

Version control

The general practice (done with Git) of tracking changes to your code over time, so you can roll back, branch, and never lose work to a bad save. Example: reverting to yesterday's version after a change breaks everything.