Skip to content

Latest commit

 

History

History
107 lines (65 loc) · 5.78 KB

how-to.md

File metadata and controls

107 lines (65 loc) · 5.78 KB
Raw

How to add a new CAPO adapter

This documentation describes the steps required to add a new price cap adapter for LST or stablecoin asset.

As a first step, contact risk providers for recommendations on capo parameters:

  • maxYearlyGrowthPercent and minimumSnapshotDelay parameters in case of an LST asset
  • fixed capo value for the stablecoin asset

Deploying a new cap adapter consists of three simple steps:

  • (optional) add a specific adapter for the new asset
  • write a deployment script
  • add and run tests

There are situations, when the sync adapter should be used together with the cap adapter. Detailed description is in the separate section.

1. Creating an adapter for LST

Depending on the asset type and data source, two options are possible:

Native exchange rate

When an LST asset adapter is added and the native contract specifies a rate:

  1. Determine which method of which contract provides the rate, and add a simplified version of the interface to the interfaces folder. For example, for weETH this is the getRate() method on the token contract, and for [osETH] this is convertToAssets(uint256 shares) on a separate vault controller.

  2. Add the specific adapter contract to the lst-adapters folder. The contract must inherit from PriceCapAdapterBase and implement only one getRatio() method.

Chainlink oracle providing rate

If there is no native contract providing an exchange rate and the Chainlink feed will be used, there is no need to add a new adapter as the generic `CLRatePriceCapAdapter should be used.

2. Deployment

As risk entities only provide growth rate and snapshot delay while snapshotRatio and snapshotTimestamp are flexible, you need to get them by yourself to use in deployment.

Get params for the deployment

  1. The first option is to use GetExchangeRatesTest:
    • Add a method to get the rate and console log for the corresponding network test. Example for osETH
    • set the block number to approximately now - minimumSnapshotDelay and run the test contract with output to the console.
  2. The second option is to do it any other way you want.

Script

Alter the appropriate deployment script:

  1. Add a function that will return the deployment code to the library. Example for weETH. The following parameters should be specified:

    • aclManager: ACL manager of he pool
    • baseAggregatorAddress: the address of the base asset feed, for ETH-based LSTs it should be ETH / USD oracle
    • ratioProviderAddress: the address of the contract, which provides the exchange ratio
    • pairDescription: description of the adapter
    • minimumSnapshotDelay: the delay provided by the risk entity, typically 7 or 14 days
    • snapshotRatio: the value of the exchange ratio X days ago
    • snapshotTimestamp: timestamp of the snapshot ratio
    • maxYearlyRatioGrowthPercent: the maximum possible annual LST growth percentage

    1.1 If the adapter is deployed on zkSync network, you'll need to create a function that only returns the encoded parameters above instead of returning the deployment code.

  2. Add the deployment script and command to the Makefile.

3. Testing

To test the adapter:

  1. Add the test to the destination network folder inside tests.

  2. Inherit it from BaseTest and implement the simple _createAdapter() method, when the specific adapter is created. Or just inherit the test from CLAdapterBaseTest.sol when Chainlink oracle is used.

    2.1. If the adapter will be tested against the zksync network:

    • add the salt parameter using the new keyword for deployment: e.g.: new CLRatePriceCapAdapter{salt: 'test'}(capAdapterParams)

  3. Specify the following test parameters:

    • adapter code
    • number of days for retrospective testing (default is 90). Check that with the specified parameters the adapter has not been capped for the last X days. A report comparing prices with the base aggregator is also generated.
    • fork parameters: network and block number
    • name for the report (something like osETH_Ethereum)

Adapter for Stablecoin

No need to add a specific adapter, existing PriceCapAdapterStable should be used.

Deployment

  1. Add a function that will return the deployment code to the library in the appropriate network's deployment script. The following parameters should be specified:

    • aclManager: the ACL manager of the pool
    • assetToUsdAggregator: the address of the asset / USD feed
    • adapterDescription: description of the adapter
    • priceCap: the value of the price cap, for example for 4% it would be int256(1.04 * 1e8)

  2. Add the deployment script and command to the Makefile.

Testing

Base test for stables would be added soon. Stay tuned.

Sync adapter

When you need to cap the asset which is not pegged to the base asset of the pool (to USD), such as agEUR, then you need to use a combination of adapters:

  1. Cap adapter for agEUR / EUR as agEUR should be capped against EUR.
  2. CLSynchronicityPriceAdapterPegToBase to combine capped adapter with EUR / USD feed to create capped agEUR / EUR / USD.

Here is the example.