With no bridge or oracle, put USDC to work on Aave V4 (Ethereum). Borrow against it on Creditcoin. The yield your deposit earns is proven across and quietly retires the debt, until you owe nothing.
Your savings make the payments. You wait and do nothing.
Repayments you make
Liquidation events
Borrow limit, earned
Proof crosses. Nothing else
How it works
You have $1,000 USDC and want cash without selling. Here is exactly what happens, in the order it happens.
It goes into RiyaEscrow, which hands it straight to Aave V4. The escrow keeps no accounting of its own. It emits one event, and that is the entire contract.
$1,000.00
collateral
A readability worker notices your deposit and asks Creditcoin to verify it. Creditcoin checks the Ethereum block itself through the Block Prover Precompile.
0x0FD2
precompile
New addresses open at a 10% limit and mint rUSD against the collateral. The limit is a credit score you have not used yet, and it increases with your usage of Riya.
$100.00
rUSD minted
Aave pays roughly 5% a year on the $1,000. Because Aave positions rebase silently, Riya periodically pulls the profit out in a real transaction, turning continuous yield into a discrete, provable fact.
≈ $50
per year
The harvest is proven the same way the deposit was. Creditcoin sees real yield arrive and knocks it off your debt. No transaction from you, no repayment schedule, no liquidation price to watch.
−$38.60
debt retired
Then you owe nothing.
$0.00
owed
You never repaid a penny. Your savings did it for you. Your savings on one chain and the loan on another.
Technical alignment
The useful test for a cross-chain design is whether it would still work, unchanged, on any other L2.
Creditcoin re-checks the Ethereum block itself through the Block Prover Precompile. Riya's ASC additionally requires receiptStatus == 1 and pins each event to the contract that may emit it. So a proof that decodes but did not succeed is rejected.
0x0FD2
Ethereum holds the money and states facts. Creditcoin decides what they mean — collateral, fee, score, limit, debt. State that exists in one place cannot drift from a copy.
no mirror state
rUSD is an ERC-20 on Creditcoin. Any Creditcoin wallet, contract or DEX can accept it. Your position never needs to leave.
rUSD · 6 decimals
Every proof travels inbound: Ethereum to Creditcoin. Riya V1 needs no outbound message, so nothing in the product is blocked on writability.
readability only
The truth: Riya's Precompile is the main reason the loan learns that the collateral earned anything.
Collateral
One asset today, and an honest reason for each of the others. Your dollars never leave Ethereum: they sit in Aave earning, and only a proof crosses to Creditcoin.
USDT
Needs its own adapter and escrow
DAI
Needs its own adapter and escrow
wETH
Needs an oracle. The ledger counts collateral in dollars
Riya’s position on Ethereum
OfflineEverything above what was supplied is yield, and harvest() may move exactly that and nothing else. Read live from the adapter.
On the rate
riya does not read an APY on chain, and does not claim one. The ledger takes the rate as an argument to selfRepayRateBps, so the assumption is the caller’s and this page owns it. Today it assumes 5.0%.
Aave V4 is deployed on Ethereum Mainnet and nowhere else, so the demo runs against a stand-in on Ethereum Sepolia. The production venue is the Aave V4 Spoke, reserve 7. Check the real rate there.
Credit score
Your borrow limit starts at 10% and climbs to 50%. It moves for one reason: yield retiring your debt. Cash repayments do not count.
Try it
19
Score
10%
Max LTV
$100
You may borrow
The credit score is relative to your cumulative debt retired.
The ladder
If repaying moved the score, anyone could borrow $100 and repay $100 on a loop without their collateral ever doing any work. Only proven yield writes to the score, so the ladder measures productive collateral.
Onboarding guide
Five steps, two chains. For V1, everything below is on testnet: no real funds are involved.
Hit Connect in the top right. Your wallet will offer to add Creditcoin Testnet if it does not have it. The same address is used on both chains. Riya reuses your Ethereum address verbatim on Creditcoin, so there is no second account to manage.
Approve the escrow, then deposit. There is a minimum deposit: small deposits cost more in gas than they are worth. Once this lands, your collateral is in Aave and earning.
You do nothing here. A readability worker picks up your deposit event, waits for the Ethereum block to be final enough to prove, and submits it to Riya's ASC on Creditcoin. Your dashboard flips from pending to credited when the ledger has it.
Your opening limit is 10% of collateral. Draw any part of it and rUSD is minted to your address on Creditcoin.
Every yield harvest is proven and applied against your debt. Nothing is asked of you. The dashboard shows the debt, the amount retired so far, and how long the rest will take at the current yield rate.
Roadmap
Phases 0 through 2 are ours to execute. Phase 3 is not.
One asset, one source chain, one destination. USDC on Ethereum Sepolia into Aave V4, harvests proven onto Creditcoin Testnet, rUSD minted against a credit score that only proven yield can increase.
Everything needed before anyone's real money is welcome.
This phase is about the Creditcoin side of the ledger becoming valuable.
These need Creditcoin features that are not released yet.
Risks
A risk you find on your own discounts everything else on the page. Here are ours, first.
Yield is measured as what Aave holds minus what we put in. If that reserve takes a loss, the collateral shrinks while the Creditcoin debt does not. Riya has no liquidation path at all, so the protocol wears the gap rather than the borrower.
What we do about it
Designed, not built. A shortfall is provable the same way yield is: the adapter marks holdings against principal on Ethereum, and Creditcoin verifies that mark through the same precompile. The fee accrued so far absorbs it, the remainder spreads across collateral, and impaired collateral stops backing new borrowing. That prices a loss the moment it is proven. It does not refund one.
A deposit is credited once its Ethereum block is final enough to prove, not the moment it lands. Expect minutes, not seconds.
Built inside a hackathon window. It has fork tests on the Ethereum contracts and no external audit. An audit is the first item in Phase 1 for exactly this reason. Do not put anything you cannot lose in front of it.
Live
Live contracts on two chains, and the Attestcoin precompiles they depend on. Every one is verifiable without trusting a word on this page.
Where every decision is made
LoanLedger
Collateral, debt, credit score, borrow limit
RiyaASC
Verifies Ethereum proofs. The only door between the chains
RiyaUSD
The borrowable dollar. 6 decimals, minted only by the ledger
Where the money sits, earning
RiyaEscrow
Custody. Emits the deposit event Creditcoin proves
AaveV4Adapter
Supplies to Aave and harvests yield back to the escrow
MockUSD
Demo dollar. Aave V4 is mainnet-only, so testnet needs a stand-in
Native to Creditcoin. Not deployed by riya
BlockProver
Verifies that an Ethereum transaction really happened
ChainInfo
Names the source chains this network can read
Documentation
Every claim on this page traces back to a contract you can read. The docs are being assembled from the build walkthrough now.
Eight checkpoints, contract by contract, with the reasoning behind each decision rather than just the code.
8 checkpoints
Deployed addresses, ABIs, events, and the exact proof format RiyaASC accepts.
Sepolia + Creditcoin Testnet
How to accept rUSD in a Creditcoin contract, and how to read a riya credit score from your own protocol.
For builders