Troubleshooting

Two things fix most problems:

• Run the same command again. Setup and deploys pick up where they stopped.
• Run pnpm everyapp status. It shows what's wrong and what to run. Check your setup with everyapp status explains every line.

Subscribe this account to Workers PaidCopy Anchor Link

Your Cloudflare account is on the free plan. Turn on Workers Paid, then run the command again. Nothing was created.

If you see Couldn't confirm this account is on Workers Paid, setup carries on. Check that you've subscribed.

Add a domain to this Cloudflare accountCopy Anchor Link

The Cloudflare account you logged in to has no domain. Get a domain, then run the command again. If you have more than one Cloudflare account, check you approved the login for the right one.

Finish activating your domain on CloudflareCopy Anchor Link

Its nameservers don't point to Cloudflare yet. Finish the nameserver step, wait for Cloudflare's email saying the domain is active, then run the command again.

Choose a workers.dev subdomain for this Cloudflare accountCopy Anchor Link

Cloudflare asks every account to pick this name once. Open the link the terminal shows, pick any name, and run the command again. Your apps never use it.

Your address already points somewhere elseCopy Anchor Link

Something on your domain already uses the address you chose. Delete that record in Cloudflare (your domain, then DNS), or choose another address.

If the record is AAAA 100:: and you saw "Choose a workers.dev subdomain" on the run before, that record is the Gateway's own: the earlier run published the address, then stopped. Since 2026-10-24, setup recognizes it and carries on. On an older copy, run git pull in your every-app folder, then the command again.

The Gateway build fails with connect ECONNREFUSED 127.0.0.1Copy Anchor Link

On Linux, setup stopped at Building the Gateway with GATEWAY_BUILD_FAILED and a line like connect ECONNREFUSED 127.0.0.1:43739. Your copy of EveryApp is older than 2026-10-24: run git pull in your every-app folder, then run the command again. If you can't update yet, run it as NODE_OPTIONS=--dns-result-order=ipv4first pnpm everyapp setup …. On Linux, localhost can mean the IPv6 address first, and the build's preview server and its own client then disagree about where to meet; newer versions of everyapp set this option themselves. Deploying an app can hit the same error, with the same fix.

Open the setup link above to create the owner accountCopy Anchor Link

Nobody opened the owner link within 15 minutes, so setup stopped waiting. The link still works for about 24 hours: open it to create your owner account. If you've lost it, run pnpm everyapp setup again for a new link. That replaces the old one.

Use the newer setup link, or get a new oneCopy Anchor Link

The owner link stopped working before anyone used it: setup ran again, maybe on another machine, and made a newer link, or the link expired. Use the newer link if you have it, or run pnpm everyapp setup again for a new one.

A new address doesn't load yetCopy Anchor Link

Cloudflare is still making its HTTPS certificate. Give it up to ten minutes.

An app in your Gateway says Needs deploy, Needs setup, Deploying or DisabledCopy Anchor Link

• Needs deploy: the app isn't online yet. Run pnpm run deploy in its folder.
• Needs setup: a key the app requires isn't set, so it won't open. Run pnpm everyapp configure my-app: it asks for what is missing. pnpm everyapp secret list my-app says which.
• Deploying: the app is still being connected. Refresh in a minute.
• Disabled: the app has been turned off.

The schema has changes no migration coversCopy Anchor Link

Your app stores something new but has no migration for it. In the app's folder, run pnpm run db:generate, then deploy again.

This checkout's Gateway is older than the live oneCopy Anchor Link

Someone updated your Gateway from a newer copy of EveryApp. In your every-app folder, run git pull and pnpm install, then try again.

Your every-app folder has a newer Gateway than the one onlineCopy Anchor Link

Not a problem. Run pnpm everyapp update to update it, or leave it: the next new app you deploy updates it too.

everyapp needs Node.js 22.18.0 or newerCopy Anchor Link

Install the latest LTS version from nodejs.org, open a new terminal window, and try again.

ERR_PNPM_LOCKFILE_CONFIG_MISMATCHCopy Anchor Link

pnpm install stopped with Cannot proceed with the frozen installation. It was told to keep the lockfile exactly as it is (--frozen-lockfile, which a CI variable in your shell turns on by itself), and the lockfile doesn't match this copy's settings.

Run pnpm install on its own, without --frozen-lockfile.

If it still fails, run pnpm --version. EveryApp is built with pnpm 10.34.5, and pnpm 10 switches to that version by itself; an older pnpm doesn't. Run npm install -g pnpm, open a new terminal window, and try again.

Command "everyapp" not foundCopy Anchor Link

You're outside your every-app folder, or haven't run pnpm install in it yet. Run cd every-app, then pnpm install, and try again.

No everyapp.config.ts in this directoryCopy Anchor Link

This command runs inside an app's folder. Run cd apps/my-app and try again.

ELIFECYCLE Command failed with exit code 2 or 130Copy Anchor Link

pnpm dev and pnpm run deploy are scripts in the app's package.json that run everyapp dev and everyapp deploy. When everyapp stops on purpose, pnpm adds this line after it. The line is pnpm's; nothing failed, and the lines above it are the message that matters.

• exit code 2: everyapp stopped for you to do something. The lines above say what, and the command to run again.
• exit code 130: you pressed Ctrl+C, and the app stopped.

pnpm everyapp dev and pnpm everyapp deploy run the same commands without the line.

Strange errors after using npxCopy Anchor Link

npx everyapp downloads an old version that doesn't work: publishing to npm is paused, and the copy on npm is from before this Gateway. A copy installed with npm install -g runs from outside your folder and says everyapp has to run from a checkout of the EveryApp repository. Always run pnpm everyapp from your every-app folder.

Still stuckCopy Anchor Link

Run the command again with --verbose at the end, and bring the output to our Discord or email it to support@everyapp.dev (everyapp.dev is ours too). Expect a fast, informal reply from a very small team. There's no status page: your Gateway runs in your Cloudflare account, so if every app is down at once, check Cloudflare's status first.