Block Analitica

Block Analitica builds risk tooling for lending protocols: dashboards that show collateral health, stress tests, and advice on which assets to list and at what limits. Its oldest client is Sky, formerly MakerDAO, where it has done risk work since 2019, and Spark, Sky's lending arm. Its site also lists Chronicle, HyperLend, Compound, Morpho, Ajna, Aave, Ethena, Badger and 3Jane.

Most of that work is advisory: it recommends and a DAO votes. Lazy Summer was different. Summer.fi's vault protocol spreads USDC across strategies it calls ARKs (a Silo market, a Term Finance market and so on), and as risk curator BA Labs set how much each ARK could take. It published a risk framework for those vaults describing how assets were classified and exposure limited. We found no public figure for what it was paid.

In both cases BA Labs had already moved to cut exposure. Its own retrospective on the July exploit, published two days later, puts the gap plainly: 'A cap of zero prevents new authorized allocations... but it does not remove an ARK from the active fleet set or from NAV accounting.' The Silo ARK had been capped for more than eight months. Removing it and freezing withdrawals were governance powers, not the curator's.

That explains the loss without clearing the curator: a risk manager that knows an asset is being wound down can push governance to remove it. Earlier, in late October 2025, BA Labs did get it right, cutting xUSD and deUSD exposure to zero before both collapsed. After July, it stepped down as curator and Summer.fi governance moved to wind the vaults down.

The November 2025 case shows the limits of a cap. BA Labs set the Silo sUSDX/USDC cap to zero on 4 November, two days before USDX lost its peg, but Stables Labs pulled its own liquidity from the market, utilisation hit 100% and the Arbitrum vault had nothing to withdraw. BA Labs' retrospective came 25 days later. A reimbursement in SUMR, Summer.fi's token, was proposed in December; we found no record that it was paid.