The Graph · Messari standardized Lending schema v3.1.0

One query. Three protocols. Three risk questions.

Unavailable

Messari's common schema gives Aave v3, Compound v3 and Spark Lend the same entities — Account, Position, Market, Token, FinancialsDailySnapshot. So the query below is written once and sent unmodified to all three. Adding a fourth protocol is one row in LENDING_SUBGRAPHS.

ETH mark—Median across protocols

Market stress—Added to the required health factor

ETH utilization—Borrowed / supplied

Liquidation intensity—7d liquidated / total borrows

Try it against a real borrower

Whose exposure?

This account is fresh, so it has no borrows anywhere — the honest answer is $0. Point the same query at a real mainnet borrower and the exposure cap trips: the policy restricts new borrowing instead of liquidating, because debt owed elsewhere is not a breach of this position's terms.

Open borrows for

—

Per deployment

Same document, every protocol

Deployment	Schema	This account	Protocol borrows	Liquidated 7d	ETH markets	ETH price

The query itself

Written once

query GalvanicProtocolRisk($account: ID!) {
  account(id: $account) {
    positions(where: { side: BORROWER, hashClosed: null }, first: 1000) {
      balance
      asset { symbol decimals lastPriceUSD }
    }
  }
  financialsDailySnapshots(first: 7, orderBy: timestamp, orderDirection: desc) {
    dailyLiquidateUSD
    totalBorrowBalanceUSD
  }
  markets(first: 200, where: { isActive: true }) {
    totalBorrowBalanceUSD
    totalDepositBalanceUSD
    inputToken { symbol lastPriceUSD }
  }
}

Not one field here is protocol-specific. USD values for positions are derived the standard Messari way — balance / 10^decimals × asset.lastPriceUSD — which works identically everywhere precisely because Token.lastPriceUSD is part of the shared schema.