Questions

Can you sign an arbitrary transaction from my relayer?

On AWS and Google Cloud, yes, and so could anyone who took over the Nonceworks server: the credential you grant can sign any digest with your key until a revocation takes effect. The relayer's policies run on our host, so they constrain the relayer process, not whoever holds the credential. Keep only gas on the relayer address, give it no authority it does not need, watch your provider's log of Sign calls, and revoke at the permission or key level if something looks wrong. On Turnkey you can enforce chain, recipient, function, value and gas rules in Turnkey's enclave, which our server cannot change as long as the user you grant is not root. The security page has the details.

What if you disappear?

Revoke the grant at the permission or key level, pull the same images by digest, and run the files you sent us on your own host. Transactions already mined are unaffected. If we are gone without notice, stop sending, wait for pending transactions to settle on chain, and start your own relayer on a fresh Redis. With notice, you get a copy of the relayer's Redis store and its storage key; your relayer loads it with one RESET_STORAGE_ON_START=true start so it uses your own config and credential.

Which chains?

Any EVM network the Relayer and Monitor support, with your RPC URLs. So far we have tested Relayer 1.8.0 on Base Sepolia and on anvil, and Monitor 1.6.0 against Base mainnet, and we test your chain before your stack goes live. We have not run the Solana or Stellar relayers.

How do upgrades roll?

Each release is staged against a copy of your configuration before your stack gets it. The operations page has the steps.

Why not KeeperHub or Tenderly?

Hosted KeeperHub provisions a Turnkey wallet for each organization, and its workflows sign with it unattended. Its self-hosted Helm chart takes your own Turnkey organization, but then you run it on your own Kubernetes. Tenderly's Web3 Actions tutorial keeps the signing key in Tenderly's secrets store. Here the key stays in your account, you can revoke the grant yourself, and you don't run the servers. If you want a dashboard and a larger company behind the service, they are the better choice.

Why not run it myself?

Many teams should, and the notes are written for that. Some teams replaced Defender with their own cron job signing through a KMS key, which works if someone on the team has time to keep it running.

Can you run our Chainlink upkeep?

If it follows the Chainlink pattern, checkUpkeep and performUpkeep, yes, once our runner is past testing. It calls your existing contract at a fixed interval from your own address, through a relayer that signs with your key, logs every run, and alerts when a run fails or when the upkeep stays due with nothing performed. The contract does not change when performUpkeep accepts any caller. So far the runner has run against Relayer 1.8.0 with a local test key on a local chain only. A Gelato resolver would need an adapter that does not exist yet.

Do you touch the Defender Actions?

Actions did not export and the Relayer has no scheduler. The Defender migration guide covers the options; a scheduler for Relayer plugins is being built.

What does the first week look like?

On the first day you send the config, we send back the validation report, you make the grant, we bring the stack up and send the address, and you fund it. For the rest of the week the stack can run alongside whatever you use now, so you can compare alerts and transactions before you switch.