Llama Lend vaults

Llama Lend is Curve’s isolated lending protocol. This integration reads V1 and V2 lender vaults.

Lenders supply the borrowed token, such as USDC, and receive ERC-4626 shares. The share value accrues borrower interest. The collateral token, such as wstETH, secures borrowers’ loans; lenders do not receive that token’s staking yield directly.

LLAMMA (Lending Liquidating Automated Market Maker Algorithm) rebalances borrower collateral through soft liquidations. This mechanism does not guarantee against losses or unavailable withdrawal liquidity.

Liquidity and utilisation

The controller holds unborrowed lending tokens. Its outstanding debt and the vault’s assets give the following metrics:

available_liquidity = max(totalAssets - controller.total_debt, 0)
utilisation = min(controller.total_debt / totalAssets, 1)

An empty vault has zero utilisation. The V1 vault includes the controller’s token balance in assets. The V2 vault deducts accrued protocol fees, so subtracting debt gives net lender liquidity in both versions. The zero separate management and performance fees reported by the adapter do not exclude protocol deductions already reflected in NAV.

Spot queries pin assets and debt to the same block. Historical scans use the same accounting through multicall and export available_liquidity and utilisation. Failed metric reads remain missing. The metadata cache version is bumped to refresh stored spot observations on the next metadata scan. Historical price rows are not automatically repaired, and the scanner retains its existing sampling policy.

Withdrawals have no scheduled waiting period, but require sufficient market liquidity and remain subject to protocol and account limits.