Restaking · EigenLayer · AVS · LRT

Restaking Security Audit

Quick Answer

  • Restaking security audits cover EigenLayer AVS slashing contracts, operator key management, liquid restaking token (LRT) security, withdrawal queue timing risks, and AVS collusion attack vectors. EigenLayer manages $15B+ in restaked ETH as of July 2026.
  • The critical risk in restaking is slashing condition correctness — if an AVS's slashing contract can be gamed (by the AVS itself or by attackers), honest operators lose capital and restakers lose delegated stake. A maliciously designed AVS is theoretically a legal theft mechanism.
  • LRT depegging risk is systemic: LRTs used as collateral in Aave or Compound can trigger cascading liquidations if a large slashing event reduces their backing ratio faster than lending protocols can process it. Free first scan, results in 60 seconds.

Restaking introduces a new risk category to Ethereum's security model: systemic slashing risk. A single large slashing event can propagate from an AVS → operator → LRT token price → lending market liquidations, affecting protocols and users far removed from the original slash. Auditing the on-chain logic that governs this cascade is the core challenge of restaking security.

Restaking Security Architecture Overview

AVS Slashing Contracts

Each AVS defines its own slashing conditions. These contracts must be deterministic, verifiable on-chain, and impossible to trigger via malicious input from the AVS operator. They are the highest-value audit target in the restaking stack.

Operator Key Isolation

Operators sign AVS attestations with ECDSA keys. Hot signing keys for frequent attestation must be isolated from admin and withdrawal keys — a single compromised key should not expose all operator stake.

LRT Token Mechanics

Liquid restaking tokens represent delegated + restaked ETH. Minting/redemption logic, oracle price feeds to lending markets, and withdrawal queue handling all introduce failure modes that can depeg LRTs during market stress.

Withdrawal Queue Security

EigenLayer's withdrawal queue creates a timing gap between initiating and completing unstaking. Slashing that occurs during this window can reduce the withdrawal amount — restakers need to understand this asymmetric risk before delegating.

Audit Your Restaking Protocol Now

AVS slashing conditions, operator key exposure, LRT redemption logic, and withdrawal queue security — free first scan.

Restaking Vulnerability Classes

VulnerabilitySeverityDescriptionExample
Slashing Condition ManipulationCriticalEigenLayer AVSs define their own slashing conditions — the on-chain logic that determines when operator stake is penalized. Maliciously designed or buggy slashing conditions can create scenarios where honest operators are slashed (losing restaker capital) or dishonest operators are never slashed (removing the security guarantee). The slashing contract is the most critical security boundary in a restaking protocol — its conditions must be verifiable, deterministic, and resistant to griefing attacks by the AVS itself.An AVS slashing contract that evaluates slashing based on off-chain signature data submitted by the AVS operator — the AVS can forge or selectively submit data to slash operators it wants to penalize while protecting its own stake.
Operator Key CompromiseCriticalRestaking operators sign attestations and participate in AVS consensus using ECDSA keys. If an operator's signing key is compromised, the attacker can: (1) sign fraudulent AVS attestations, triggering valid slashing of the operator's restaked capital, (2) perform double-signing or other slashable offenses to drain delegated stake, or (3) participate in coordinated attacks on the AVS's security guarantee itself. Key compromise is the highest-probability path to operator capital loss in the current restaking ecosystem.An EigenLayer operator using a single hot signing key for all AVS attestations — one server compromise means the attacker can double-sign on any AVS the operator validates, triggering slashing for every delegator who chose that operator.
LRT (Liquid Restaking Token) DepeggingHighLiquid restaking tokens like eETH (ether.fi), rsETH (Kelp), and ezETH (Renzo) represent staked + restaked ETH positions as tradeable tokens. These tokens rely on the underlying ETH being redeemable at approximately 1:1 with the represented value. If a large slashing event reduces the underlying ETH value faster than the LRT price adjusts, or if withdrawal queues prevent timely redemption during market stress, LRT holders can be left holding tokens worth significantly less than par value.A 5% slashing event across a large operator set reduces the ETH backing ratio of an LRT from 1:1 to 0.95:1. If the LRT is used as collateral in lending protocols (Aave, Compound), positions that were at 80% LTV are suddenly liquidatable — a cascading liquidation event across all lending markets that accepted the LRT.
Withdrawal Queue Timing AttackHighEigenLayer imposes withdrawal queuing periods that delay when operators can withdraw staked ETH. These delays create a timing asymmetry: if an operator discovers a slashable condition before the slashing is enforced, they can initiate withdrawal to move capital out of the slashable queue before the slash is applied. Conversely, a sudden slashing event during the withdrawal queue period can reduce a restaker's capital mid-withdrawal — the final withdrawal amount is less than the amount queued.An operator detects that a bug in an AVS slashing contract will cause spurious slashing. The operator initiates withdrawal before the slashing is applied, but the withdrawal queue is 7 days. If EigenLayer processes the slash before the 7-day withdrawal completes, the operator loses stake they thought they'd already moved to safety.
AVS Collusion for Invalid SlashingHighEigenLayer's security model assumes AVSs act honestly and only slash provably misbehaving operators. An AVS controlled by a malicious actor could design slashing conditions that allow it to slash honest operators under contrived pretenses — effectively using the AVS framework to steal delegator capital without a real security violation. Restakers who choose operators registered with a malicious AVS are exposed to this attack.An AVS whose slashing contract allows the AVS admin to submit any 'violation proof' signed by the AVS's own key, with no on-chain verification that the proof represents actual operator misbehavior — the AVS admin slashes all operators, stealing delegator stake.

Restaking Security Audit — Frequently Asked Questions

What is a restaking security audit, and what does it cover?
A restaking security audit reviews the smart contracts that enable Ethereum validators and ETH stakers to secure additional protocols (AVSs) using their already-staked ETH. The audit covers: (1) EigenLayer integration contracts — how operator registration, delegation, and withdrawal are handled; (2) AVS slashing contracts — the conditions under which operator stake is penalized and whether those conditions can be gamed; (3) Liquid restaking token (LRT) contracts — the minting, redemption, and price stability mechanisms for protocols like ether.fi, Kelp, and Renzo; (4) Operator key management — signing key exposure and multi-key isolation. As of July 2026, EigenLayer manages over $15B in restaked ETH, making it one of the highest-value audit targets in DeFi.
How does EigenLayer restaking work, and where are the key security risks?
EigenLayer allows ETH stakers to 'restake' their staked ETH — simultaneously securing Ethereum PoS AND additional protocols called Actively Validated Services (AVSs). The staked ETH serves as economic collateral for both: a validator who misbehaves on Ethereum OR violates an AVS's slashing conditions can lose a portion of their restaked capital. The key security risks are: (1) AVS slashing condition correctness — whether the conditions are valid and resistant to gaming, (2) Operator key security — whether signing keys are isolated from admin infrastructure, (3) Cascading slash risk — whether a slashing event in one AVS triggers unexpected liquidations in LRT-backed lending markets, and (4) Withdrawal timing — whether the queue delay creates asymmetric risk for operators who detect slashable conditions.
What is a Liquid Restaking Token (LRT) and what are its security risks?
Liquid restaking tokens (LRTs) are ERC-20 tokens issued by protocols like ether.fi (eETH), Kelp DAO (rsETH), and Renzo (ezETH) that represent a claim on restaked ETH. Holders earn both Ethereum staking yield (~3-4%) and AVS yield without managing validator infrastructure. Security risks: (1) Slashing pass-through — a large slashing event reduces the ETH backing ratio below par value, depegging the LRT; (2) Withdrawal delay — LRT redemption requires queueing unstaking from EigenLayer, creating a timing gap where the LRT may trade at a discount during market stress; (3) Oracle dependency — lending protocols price LRTs via price oracles, and stale or manipulated LRT price data creates bad debt in lending markets; (4) Admin key custody — LRT contract admin functions that control fee parameters or reward allocation represent centralized risk.
Can an AVS steal restaker funds through its slashing contract?
In theory: yes, if the slashing contract is designed maliciously and restakers don't verify AVS contract code before delegating to operators that service that AVS. EigenLayer's slashing model assumes AVS operators are economically rational and honest — an AVS that slashes honest operators loses credibility and its ability to attract future operators. But 'reputation loss' is not a reliable deterrent when an AVS can slash once and exit with stolen funds. The protection for restakers: only delegate to operators registered with AVSs whose slashing contracts have been audited and whose slash conditions are deterministic, verifiable on-chain, and immune to admin override.
How much does a restaking protocol security audit cost?
A restaking audit cost scales with protocol complexity. A single AVS integration (operator registry + slashing contract): $20,000–$50,000. A full LRT protocol with EigenLayer integration, token minting/redemption, fee management, and withdrawal queue handling: $80,000–$250,000. Full-stack audits of large LRT protocols (ether.fi, Kelp) involve multiple audit firms and multiple rounds. Protocols with novel slashing condition designs — custom verification logic, ZK proof-based attestation, BLS signature verification — require specialized expertise and command higher rates. SmartContractAuditor.ai provides instant analysis for slashing condition logic, operator key isolation patterns, LRT redemption reentrancy, and withdrawal timing vulnerabilities — free for the first scan.
Written by Duron Epps, Founder of SmartContractAuditor.ai · Last updated July 2026