In order to optimize the cost of your recurring predicate executions, we recommend first deploying your predicate. This can be done using the Fuels CLI and running the deploy command .
By deploying the predicate, its bytecode is stored on chain as a blob. The SDK will then produce bytecode that can load the blob on demand to execute the original predicate. This far reduces the repeat execution cost of the predicate.
To deploy a predicate, we can use the Fuels CLI and execute the deploy command .
This will perform the following actions:
forc version We can then utilize the above generated types like so:
import { Provider, Wallet } from 'fuels';
import { LOCAL_NETWORK_URL, WALLET_PVT_KEY } from '../env';
import { ConfigurablePin, ConfigurablePinLoader } from '../typegend/predicates';
const provider = await Provider.create(LOCAL_NETWORK_URL);
const wallet = Wallet.fromPrivateKey(WALLET_PVT_KEY, provider);
const receiver = Wallet.generate({ provider });
const baseAssetId = provider.getBaseAssetId();
// We can deploy dynamically or via `fuels deploy`
const originalPredicate = new ConfigurablePin({
provider,
});
const { waitForResult: waitForDeploy } = await originalPredicate.deploy(wallet);
await waitForDeploy();
// First, we will need to instantiate the script via it's loader bytecode.
// This can be imported from the typegen outputs that were created on `fuels deploy`.
// Then we can use the predicate as we would normally, such as overriding the configurables.
const loaderPredicate = new ConfigurablePinLoader({
data: [23],
provider,
configurableConstants: {
PIN: 23,
},
});
// Now, let's fund the predicate
const fundTx = await wallet.transfer(
loaderPredicate.address,
100_000,
baseAssetId
);
await fundTx.waitForResult();
// Then we'll execute the transfer and validate the predicate
const transferTx = await loaderPredicate.transfer(
receiver.address,
1000,
baseAssetId
);
const { isStatusSuccess } = await transferTx.waitForResult();