Lazy Summer Protocol

Rejected venues wait the longest for re-review; a rejection has to earn another look before the scheduled date.

Lazy Summer Protocol is a yield aggregator on Ethereum, Base, and Arbitrum whose automated keepers rebalance deposits across underlying DeFi venues inside risk parameters set through protocol governance. At the 2026-08-16 survey the tracked record held only about $1,100 on Arbitrum, with zero current samples on its other listed chains. Below that floor, the book one practice would move on the same research becomes the exit crush, so the file is rejected on size, whatever the protocol’s quality. At scale it would face the delegated-allocation problem: an automated allocator chooses the venues, which is the decision we are paid to make.

Kill criteria: the events that revoke this verdict
• TVL sustained above $100M for 30 days
• Publishes allocator mandates, curator accountability, and per-vault disclosure that let the delegation be underwritten

The research file

Mechanism applicability

Lazy Summer depositors receive shares in Lazy Vault Fleets. A FleetCommander allocates assets among ARKs that represent lending, staking or other DeFi strategies; RAFT harvests and swaps rewards back into the deposit asset. Keeper agents move capital among ARKs within governance-set limits. This is delegated multi-venue allocation, but the tracked system is currently so small that the shared version-1 below-materiality dossier decides first.

Current observation and lifecycle

The DefiLlama protocol API read on 2026-08-16 classified Lazy Summer as a Yield Aggregator and reported approximately $1,100 on Arbitrum, with zero current samples on Base, Ethereum, Hyperliquid L1 and Sonic. Current Summer documentation still describes active Lazy Vault, ARK, governance and Staking V2 systems. The registry preserves the established Ethereum, Base and Arbitrum product perimeter while expressly recording that only Arbitrum had a nonzero surveyed balance.

Control and exit applicability

Governance approves Fleets, ARKs, constraints and initially whitelisted keepers; offchain monitoring selects rebalances within FleetCommander limits. Depositors inherit every selected strategy, reward-swap and keeper dependency plus a 1% AUM fee for most vaults. A Buffer ARK supports ordinary withdrawals, while a force withdrawal must unwind deployed ARKs when the buffer is insufficient, so stressed exit depends on underlying venue liquidity and gas.

Why the class rule decides

At roughly $1,100 of currently attributable TVL, even the minimum $1M advised allocation would exceed the measured system by about 900 times. The shared version-1 below-materiality dossier therefore decides before delegated-allocation underwriting. Reopen after reconciled protocol TVL remains above $100M for 30 consecutive days, then review every Fleet and ARK, curator and keeper accountability, governance and bridge controls, incidents, fees, proposed-size force withdrawals, legal access and named simpler single-venue alternatives.

Sources

The claims above trace to these. Where a number could not be independently verified, the thesis says so.

• Summer — Lazy Vault and Fleet architecture · primary · accessed 2026-08-16
  Supports: Lazy Vault, FleetCommander, ARKs, RAFT, vault shares, strategy types
• Summer — ARK allocation and unwind mechanism · primary · accessed 2026-08-16
  Supports: ARK, allocation, board, disembark, harvest, Buffer ARK
• Summer — keeper and rebalancer controls · primary · accessed 2026-08-16
  Supports: keeper, whitelist, governance constraints, rebalancing, offchain monitoring
• Summer — buffer and force withdrawals · primary · accessed 2026-08-16
  Supports: withdrawal buffer, force withdrawal, underlying ARK, FleetCommander, exit liquidity
• DefiLlama — Lazy Summer Protocol survey record · secondary · accessed 2026-08-16
  Supports: current TVL, Arbitrum, zero chain samples, Yield Aggregator category, survey observation

Inherited controls

The verdict above grades the protocol layer. Every position also inherits the asset it holds and the chain it settles on. The least safe layer sets the position’s grade, and the position table names which one that is.