More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    EAS Simulator: Cloud iOS and Android Testing

    1

    Run and verify an Expo app on a cloud iOS simulator or Android emulator from any machine: access check, the right build for the goal, drive it, capture proof, clean up.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    EAS Simulator: Cloud iOS and Android Testing

    EAS Simulator: Cloud iOS and Android Testing

    Example session with this skill installed

    I'm developing on a Linux machine with no Mac access. I need to check my Expo app's login flow on iOS and attach screenshots to a pull request. How do I actually do this?

    • Read your context and instructions
    • Compiled the eas simulator

    SESSION RUN — Login flow check on iOS, from a Linux host.

    1. ACCESS FIRST: npx --yes eas-cli@latest simulator:availability --json returns available. EXPO_TOKEN is set in the environment (no browser login exists on this host) and whoami confirms the account. If availability had said no, the run stops here and falls back instead of retrying starts.

    2. START, NAMED: simulator:start --platform ios --type agent-device --non-interactive --name "Login flow PR screenshots" --max-duration-minutes 30. simulator:get --json shows IN_PROGRESS before anything touches the device.

    3. BUILD MATCHES THE GOAL: This is a static check of current code, so a fresh local release build from current source is installed — not the build already sitting on the simulator, which predates the login change and would screenshot the old flow while looking current.

    4. DRIVE IT: Through simulator:exec, the app opens by bundle ID. snapshot -i returns the accessibility tree; the email field is @e3, password @e5, Sign in @e7. Fill both fields, press Sign in, wait for the home screen snapshot, then screenshot ./login-success.png. A wrong-password pass runs too: record start ./login-error.mp4, submit bad credentials, record stop — the error shake is motion, so it is proven on video, not a still.

    5. CLEAN UP: simulator:stop, then the session file is reset so the next run does not inherit a dead session ID. Screenshots and the clip are attached to the PR with the build fingerprint noted, so reviewers know exactly which code the images show.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Your development machine is Linux, a CI runner, or a cloud sandbox — and the change you just made needs to be seen running on iOS before anyone trusts it. EAS Simulator hosts the device remotely, but the tooling is experimental, the sessions are short-lived and billed, and the failure modes are quiet ones: a stale release build that makes you verify yesterday's code, an idle session burning budget, a session token leaked into a log. This skill runs the whole lifecycle with the traps marked. It starts with the cloud-vs-local decision, made fresh every time, then the read-only availability check that stops you before a doomed start. The session shape is fixed: check access, start named and time-boxed, confirm it is live, install, drive, stop. Build selection gets its own section because it is the number one failure mode — a dev build for live iteration, a fresh build from current source for a static check, a named artifact only when that artifact is the question. Driving the device goes through the exec bridge and accessibility snapshots, with screenshots for state and recordings for anything that moves. A failure playbook covers the dead daemon, the stale session file, and the Metro port conflict, with a reset rule that stops you poking live state. Cleanup is part of the job: session stopped, session file cleared, nothing left running.

    • EAS Simulator is experimental and limited-access; if your Expo account lacks access, the skill's availability check stops the run and falls back to local options.
    • A simulator cannot prove real-device behavior: push delivery, hardware biometrics, carrier networks, thermals, or representative performance.
    • Session commands and flags change between eas-cli versions; the skill treats the installed CLI's help as authoritative over any remembered syntax.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 24 days ago

    Requires

    macOS
    eas-cli
    Node.js
    agent-device

    Permissions

    Terminal / Shell
    Listed24 days ago
    Updated21 days ago

    What's inside

    Frequently Asked Questions

    Usually no. On a Mac the local simulator is the default; cloud earns its place for an iOS version you lack, a shareable session, or an agent-driven run. The skill's first section makes that call explicitly every time.

    Almost always the build: a leftover release build cannot live-reload, so Metro reconnects and the device silently shows stale code. The fix is to install a dev build for live work — the skill covers this as the number one failure mode, plus the one-clean-reset rule when a session goes sideways.

    It proves the flows you drove, on that build, in a simulator. It does not prove push delivery, biometrics hardware, real network behavior, or performance. The skill names what a session did not prove as part of its verdict.

    New in Mobile Development