Skip to main content
Bridge transfers can encounter two error types. Hard errors stop execution. Soft errors let you recover and retry the bridge transfer. The App Kit SDK provides error handling that helps you respond in both cases. Hard errors throw exceptions such as validation errors, configuration issues, and authentication problems. Soft errors occur during the bridge transfer but return enough transaction information for recovery. Examples include insufficient balance, network timeouts, and RPC connectivity issues. This guide helps you identify failure points, recover partial bridge transfers, and implement error handling patterns.
The examples below are focused recovery snippets, not complete runnable scripts. They assume you have already configured kit, sourceAdapter, and destinationAdapter. See Adapter setups for setup options.

​
Bridge transfer failures

This section explains how to identify where a bridge transfer failed and resume the bridge transfer manually.

​
Transaction steps overview

Each bridge transfer uses Circle’s CCTP protocol provider, which breaks each transaction into various steps:
  • approve: Allows the contract to spend USDC.
  • burn: Burns USDC on the source blockchain and generates an attestation.
  • fetchAttestation: Waits for Circle to sign the burn proof.
  • mint: Mints USDC on the destination blockchain with the attestation.

​
Bridge result details

When a bridge transfer fails, the App Kit SDK returns a BridgeResult object showing which steps completed and which failed. Use this result to decide whether to retry the transfer, inspect a transaction on a block explorer, or surface a recoverable state to your application. Focus on these BridgeResult properties during recovery:
  • result.state - shows whether the bridge transfer succeeded or failed (pending, success, error)
  • result.steps - each object contains:
    • name: the name of the step
    • state: the status of the step
    • txHash: the transaction hash if the step completed
    • error: an error message if the step failed
This example shows a returned result object for a transaction that failed when fetching an attestation:
Shell
result.state: 'error'
result.steps: [
  { name: 'approve', state: 'success', txHash: '0x123...' },
  { name: 'burn', state: 'success', txHash: '0x456...' },
  { name: 'fetchAttestation', state: 'error', error: 'Network timeout' },
]

​
Step analysis

This example shows how to check for completed steps and use a helper function to find specific steps:
TypeScript
// Start a bridge transfer that might fail
const result = await kit.bridge({
  from: { adapter: sourceAdapter, chain: "Ethereum_Sepolia" },
  to: { adapter: destinationAdapter, chain: "Arc_Testnet" },
  amount: "1.00",
});

// Check which steps completed successfully
console.log("Bridge transfer state:", result.state);
console.log("Steps:", result.steps);

// Helper function to find specific steps
const getStep = (stepName: string) =>
  result.steps.find((step) => step.name === stepName);
const approveStep = getStep("approve");
const burnStep = getStep("burn");
const attestationStep = getStep("fetchAttestation");
const mintStep = getStep("mint");

​
Recovery scenarios

This section describes how you can implement recovery patterns.

​
Retry a failed bridge transfer

If a bridge transfer fails, you can retry it with the retry method. Pass the failed BridgeResult and the to and from adapters. This example shows how the retry method works:
TypeScript
const result = await kit.bridge({
  from: { adapter: sourceAdapter, chain: "Ethereum_Sepolia" },
  to: { adapter: destinationAdapter, chain: "Arc_Testnet" },
  amount: "1.00",
});

if (result.state === "error") {
  const retryResult = await kit.retry(result, {
    from: sourceAdapter,
    to: destinationAdapter,
  });
  console.dir(retryResult, { depth: null, colors: true });
} else {
  console.dir(result, { depth: null, colors: true });
}

​
Retry after a failed mint step

This pattern shows how to retry when the mint step fails. The failingDestinationAdapter placeholder represents a bad destination signer or RPC setup used to exercise the retry path:
TypeScript
import type { BridgeResult } from "@circle-fin/app-kit";

const findErrorStep = (result: BridgeResult) => {
  if (result.state === "error") {
    return result.steps.find((step) => step.state === "error");
  }
  return null;
};

const result = await kit.bridge({
  from: { adapter: sourceAdapter, chain: "Ethereum_Sepolia" },
  to: { adapter: failingDestinationAdapter, chain: "Arc_Testnet" },
  amount: "1.00",
});

console.log("INITIAL RESULT");
console.dir(result, { depth: null, colors: true });

if (result.state === "error") {
  const errorStep = findErrorStep(result);
  if (
    errorStep &&
    errorStep.errorMessage?.includes("gas required exceeds allowance") // This is an example error message
  ) {
    const retryResult = await kit.retry(result, {
      from: sourceAdapter,
      to: destinationAdapter,
    });
    console.log("RETRY RESULT");
    console.dir(retryResult, { depth: null, colors: true });
  }
}

​
Common issues

This section lists common issues and solutions.

​
Insufficient balance

Ensure you have enough USDC in your wallet before a bridge transfer to avoid an insufficient balance error. This EVM example checks your wallet balance:
TypeScript
import { formatUnits } from "viem";

const balanceAction = await sourceAdapter.prepareAction(
  "usdc.balanceOf",
  {},
  { chain: "Arc_Testnet" },
);
const balance = await balanceAction.execute();
console.log(`USDC balance: ${formatUnits(BigInt(balance), 6)}`);

​
Transaction stuck or failed

If a transaction is stuck or failed, check the transaction on a block explorer with the returned txHash. For Solana bridge transfers, use Solana Explorer or SolScan. If the transaction failed during the bridge transfer, check the returned result.steps to see which transaction steps completed.

​
Best practices

Follow these practices for prevention, recovery, and monitoring to improve reliability. Prevention
  • Test your integration on testnets before deploying on mainnet.
  • Monitor gas prices and adjust during network congestion.
  • Use dedicated RPC providers such as Alchemy or QuickNode.
  • Implement multiple RPC fallbacks.
  • Wrap all bridge transfers in try-catch including adapter setup and bridge calls.
Recovery
  • Always save the bridge transfer state for recovery scenarios.
  • Verify which steps completed before attempting recovery.
  • Use appropriate timeouts and give network operations enough time to complete.
  • Implement exponential backoff and use increasing delays for retry logic.
Monitoring and debugging
  • Use block explorers to verify transaction status.
  • Save intermediate results and persist bridge transfer state for recovery scenarios.