AI agents and Expo overview

Edit page

Build and publish Expo and React Native apps with AI coding agents such as Claude Code, Codex, and Cursor.

Claude Code, Codex, Cursor, and other AI coding agents can help you build, upgrade, debug, and deploy your Expo and React Native projects. When you use create-expo-app to create a new project, that new project is already set up with the configuration files required by an AI agent. This setup is also committed to your project, so everyone on your team starts from the same SDK version and project instructions, and their agents read the same project context.

How Expo supports AI agents

Three pieces work together to give an AI agent consistent, Expo-specific context:

Expo Skills: A plugin that adds Expo-specific instructions and slash commands to the agent. The agent applies known-good Expo patterns (SDK upgrades, EAS Workflows, native UI with Jetpack Compose and SwiftUI, API routes) instead of guessing from training data. Skills install once per machine.
Expo MCP Server: A remote Model Context Protocol (MCP) server that gives the agent live access to the latest Expo documentation, EAS Build history, EAS Update channels, and TestFlight metadata. The agent can install SDK-matching packages, read build logs, and take simulator screenshots through it.
Project context files: create-expo-app writes AGENTS.md, CLAUDE.md, and .claude/settings.json at the project root. These are the first thing the agent reads when you open the project. They point the agent to the documentation for the Expo SDK version your project targets.

Set up Expo Skills and Expo MCP Server

Expo Skills and the Expo MCP Server work with every supported agent. Set up each one by following the guides below:

Expo Skills

Install the plugin that teaches agents known-good Expo patterns, and browse the full list of available skills.

Expo MCP Server

Connect the remote Expo MCP Server to give agents live access to Expo documentation and EAS.

Pick an agent

Each per-agent guide covers install, setup, and example prompts for that agent:

Claude Code

Set up Anthropic's terminal agent with Expo, including project context and example prompts.

Codex

Set up OpenAI's terminal agent with Expo, including project context and example prompts.

Cursor

Set up the AI-first code editor with Expo, including project context and example prompts.

Project context files for agents

Each agent looks for project context in a different place. Instead of forcing one convention, create-expo-app CLI adds the following configuration files to a project's root when you create a new Expo project using the CLI:

File	Read by	Purpose
AGENTS.md	Codex and Cursor directly. Claude Code through CLAUDE.md.	Points the agent to the Expo documentation matching your project's SDK. This is the source of truth for project-level instructions.
CLAUDE.md	Claude Code on startup.	Contains `@AGENTS.md`, which imports AGENTS.md into Claude Code's context.
.claude/settings.json	Claude Code on startup.	Pre-enables the official Expo plugin from the Claude Code plugin marketplace.

Each file targets a different agent convention, so the same project works with Claude Code, Codex, or Cursor without per-agent configuration.

Verify the setup

To confirm an agent can read your project, open a session for your AI agent within your Expo project and run this prompt:

Example prompt

Open package.json and tell me which Expo SDK version this project targets.

If the agent replies with the SDK version from package.json, the agent is reading your project correctly.