1. Abstract

2. Protocol Workflow

3. Risk Scenarios

4. Cap Risk Profile

   1. AVS Risk Dimensions

   2. Credit-Based Cryptoeconomic Security

5. Stakeholder Interactions: Cap Across the Stack

6. Conclusion

Disclaimer:

The term “AVS” (Actively Validated Service) will be used here to refer to Cap as any protocol (AVS, Network, BSN, etc.) that leverages restaked collateral from restaking marketplaces to enforce its validation needs. The term "SSP" (Shared Security Protocols) will be used here to represent the restaking marketplaces (EigenLayer, Symbiotic, Babylon, SatLayer, etc.) that aggregate this demand and supply security and validation to AVSs.

The Covered Agent Protocol (CAP) is an AVS protocol that extends restaking to credit-backed stablecoin issuance, replacing traditional infra and consensus validation risks with credit, market, and counterparty exposure. It issues two assets: cUSD, a fully redeemable dollar-pegged stablecoin backed by a reserve of blue-chip stablecoins; and stcUSD, a yield-bearing savings instrument created by staking cUSD. Accredited financial institutions (Operators) borrow these stablecoins against restaker-pledged collateral and deploy them in pre-approved DeFi strategies.

Restakers earn a fixed rate negotiated with operators, while stcUSD holders earn a variable, stable-denominated fee. Operators retain any surplus yield after repayment as compensation for strategy execution. The protocol is underpinned by deterministic slashing and legal enforceability: repayment defaults or collateral-health breaches trigger automated liquidation and redistribution of collateral to the reserve, with legal recourse ensuring restakers are covered, thereby preserving cUSD’s peg integrity.

This risk report evaluates Cap’s risk profile, covering its architecture, dependencies, and enforcement mechanisms. It analyzes capital efficiency, liquidity constraints, and incentive alignment between restakers and operators, and examines how credit and market events can propagate through the protocol. We also model its credit-backed cryptoeconomic security framework and detail the system-level interactions between Cap and its key stakeholders.

Cap’s slashing-secured lending architecture enables restakers to act as underwriters for stablecoin loans issued to whitelisted regulated financial institutions, such as banks, market makers, HFT firms, or private equity funds. Unlike traditional DeFi lending markets where borrowers post their own collateral, Cap inverts the model: restakers pledge collateral, allowing whitelisted operators to borrow directly from Cap’s reserve pool of whitelisted stablecoins. In return, restakers earn a fixed restaking fee, while the protocol enforces safety through liquidation thresholds and slashing mechanisms, through restaking marketplaces, tied to objective health metrics.

Each loan is structured as an overcollateralized, one-to-one credit relationship between a restaker and an operator (agent), governed and enforced by off-chain legal agreements, and its repayment directly accrues to the stcUSD yield stream. These agreements define collateral amounts, loan duration, restaking fee (interest), maximum borrow limits, among other relevant terms. Cap overlays this relationship with purely on-chain enforcement logic—setting protocol-wide liquidation thresholds (akin to Aave), and configuring borrow interest rates based on a base rate plus a utilization-based spread. These borrow rates directly translate into the stablecoin yield paid to stcUSD holders.

Restakers not only accept slashing risk, but also assume responsibility for counterparty evaluation, ongoing monitoring, and agreement enforcement. The result is a high-agency, high-responsibility model that ties yield more explicitly to trust and underwriting performance than in passive AVS architectures.

The workflow can be divided into the following stages:

• User mints cUSD: A user deposits ~$1 worth of a trusted, Cap-whitelisted stablecoin (e.g., USDC, USDT) into Cap’s reserve vault, receiving 1 cUSD in return—a fungible token redeemable against the pool.

• User stakes cUSD to earn yield: Users stake cUSD into the stcUSD contract, receiving transferable and fungible stcUSD that accrues protocol fees over time and is redeemable only by the original depositor.

• Restaker delegates stake to operator: A restaker selects a specific operator and delegates restaked collateral (via Symbiotic or EigenLayer), creating a dedicated credit line that can be borrowed against up to a restaker-defined maximum.

• Off-chain credit agreement signed: The restaker and operator execute a bilateral legal agreement defining loan terms, restaking fee, max borrow amount, and repayment obligations.

• Operator borrows stablecoin from reserve: Once delegation is active and the agreement confirmed, the operator borrows stablecoins from Cap’s reserve and deploys them in yield strategies. Cap sets the borrow interest rate, which compounds over time and contributes to stcUSD yield.

• Yield strategy executed successfully: The operator earns yield using borrowed capital (e.g., through arbitrage, basis trades, or credit strategies) and prepares for repayment.

• Operator repays loan + interest + restaking fee: At maturity, the operator returns the borrowed stablecoin to the reserve, pays interest to the protocol (which accrues to stcUSD holders), and the restaking fee to the underwriting restaker.

• Fees distributed to restakers: Restaking fees accrue and can be claimed periodically by the restaker or upon loan completion, paid in the same asset as the borrowed principal.

• Reserve fully replenished: Loan repayment refills the reserve in full (similarly to Aave USDC deposit pools). stcUSD remains redeemable for cUSD at any time, subject to any queueing mechanics.

• Operator retains excess yield: Any surplus yield (after covering borrow interest and restaker fees) remaining after repayment is kept by the operator as compensation for strategy execution and risk-taking.

• Health factor falls below threshold: Loan safety deteriorates due to collateral value drop, strategy underperformance, or missed repayments.

• Grace period initiated: A short window allows the operator to repay the full loan or the restaker to increase delegation to restore solvency. If neither act, liquidation proceeds.

• Liquidation event triggered: Cap initiates a Dutch auction mechanism to alleviate the operator’s debt, where liquidators cover the debt using the restaker’s slashed funds.

• Slashing executed: The slashed funds are sold to liquidators at a discount, serving as the liquidation bonus incentive (which is larger for faster bids).

• Funds redistributed to reserve: The stablecoin provided by liquidators is routed back into Cap’s reserve pool, ensuring full coverage of the restaker’s debt and protecting stcUSD holders from losses.

• Redistribution to restaker, operator offboarded: The restaker receives their funds directly from the respective operator, by enforcing their legal agreement. The operator is removed from eligibility, and the restaker reassesses future delegations.

The dual-path ensures that operators remain accountable for the capital they deploy while enabling restakers to earn yield tied to real-world counterparty risk. Cap’s workflow is inspired by protocol parameters from Aave-style lending practices (e.g., health factors, liquidation thresholds, etc.) but introduces a novel trust-based underwriting marketplace—where productive delegation replaces passive validation, and risk is tightly scoped to bilateral agreements.

Cap shifts restakers from passive security providers to active credit underwriters, with delegated stake directly exposed to operator-specific credit, execution, and liquidation outcomes. This transforms slashing from a probabilistic penalty on missed validator duties into a deterministic, principal-sized loss tied to loan performance.

Two objective health-factor breaches trigger enforcement:

1. Collateral value falling below protocol thresholds;

2. Failure to meet repayment obligations.

Both lead to Dutch auction liquidations to restore solvency. These are exogenous, credit-driven triggers, moving the risk scope away from cryptographic correctness and into market volatility, counterparty solvency, and credit execution.

Cap operates within a layered dependency stack: execution and liveness layer (Ethereum), restaking middleware (EigenLayer, Symbiotic), oracles (Chainlink), bridges (LayerZero), and DeFi settings (e.g credit health and collateral liquidity). Its risk surface extends beyond the direct restaker–operator relationship. Failures in any upstream component can impair liquidations, liveness, or collateral recovery. In parallel, correlated strategy choices or market-wide liquidity shocks can trigger multiple operator defaults at once, amplifying the slashing impact across the system.

• High Slashable Ceiling
  Cap differs from most AVSs in that slashing is fully deterministic and can extend up to the operator’s full outstanding debt at default. Instead of a partial slashing, the result can be an immediate loss of a large portion, or even all, of the delegated stake, depending on the credit shortfall. From a VaR perspective, this produces a nontrivial tail-loss profile where allocation size directly determines loss severity. Because slashing is triggered by repayment failure rather than on-chain misconduct, common containment mechanisms in validator-based AVSs offer limited protection. Effective mitigation relies on a different set of parameters: prudent liquidation thresholds, exposure limits, and diversification across uncorrelated credit lines.

• Liquidation-Linked Slashing
  Every operator position is continuously monitored via its health factor (debt-to-collateral ratio). Breaches trigger a short grace period, followed by collateral liquidation and, if undercollateralized, slashing of the delegated stake. Because the trigger is mechanical, even brief drawdowns, liquidity gaps, or momentary execution failures can crystallize losses for restakers. This coupling of slashing to liquidation mechanics creates a tight feedback loop between market volatility and credit enforcement, where adverse price action or auction underperformance can convert unrealized risk into permanent capital loss.

• Restaked Capital Lock-In Risk
  Delegated collateral is contractually locked for the loan term plus any withdrawal latency, creating a fixed illiquidity period that cannot be shortened without forfeiting fees. This structural constraint limits restaker flexibility to reprice risk, reallocate to safer operators, or respond to macro shocks until the position reaches maturity.

• Covariance Risk in DeFi Strategies
  While designed for per-operator isolation, risk can synchronize if multiple operators crowd into the same high-risk venues or strategies (e.g., a failing bridge, overleveraged derivatives pool, thin-liquidity MEV arb). Such event would produce narrowly-correlated slashing at the DeFi—not restaking—layer, where multiple defaults emerge from a single market shock. Mimetic strategy adoption, where operators chase short-lived high yields, only amplifies this effect.

• Stablecoins Depeg Risk (incl. cUSD)
  While loans are issued in trusted stablecoins, the system remains exposed to depeg events. Loans in USDC or USDT inherit the redemption, liquidity, and solvency risks of those assets. A severe or prolonged depeg during liquidations could depress auction proceeds, impair reserve recapitalization, and trigger redemption delays. If cUSD itself trades at a discount—whether from bridge failures, liquidity drains, or loss of confidence—the impact compounds, potentially stressing both restaker recoveries and stablecoin holder redemptions.

• Liquidity Risk
  Even after a position unlocks, practical access to funds can be delayed by market and infrastructure factors. Liquidation-related auction mechanics, collateral market depth, and dependencies on upstream systems (e.g., SSP withdrawal epochs) can slow payout or capital rotation. These constraints are situational and variable, requiring real-time liquidity risk models.

• Price Volatility Risk
  ETH or LST collateral drawdowns can trigger liquidations and slashing purely from adverse price action, without any operator misconduct, closely mirroring CDP dynamics. Rapid market swings—especially in thin LST liquidity conditions—can exacerbate undercollateralization risk, forcing fire-sale auctions that crystallize losses for restakers.

• Slashing Execution Risk
  Cap’s slashing is executed on Ethereum through EigenLayer or Symbiotic, which route restaker delegations and trigger enforcement when an operator defaults, while collateral liquidations are handled internally via Cap’s auction system. Both systems provide deterministic execution, transparent governance hooks, and strong settlement guarantees, supporting reliable fault enforcement. It is critical to closely assess the slashing process efficacy of SSPs like EigenLayer and Symbiotic—our own research Restaking Protocols Infra Risk Framework V2 offers a structured basis for such evaluation. However, failure or delays in auction execution, withdrawals, oracle communication, or stake redistribution could impair enforcement, leaving restakers temporarily or permanently undercompensated.

• Active Risk Monitoring Requirements
  Cap’s design assumes restakers (or LRT vault managers) will actively monitor operator credit health, tracking LTV ratios, yield performance, repayment discipline, and macro risk factors. This represents a material shift from pooled-validator models where monitoring is collective and risk is diffuse; in Cap, inattentive oversight can allow deterioration to cross liquidation thresholds before corrective action is possible.

• Parameter Governance Uncertainty
  Key enforcement variables, such as liquidation bonus percentage, grace period duration, and liquidation window length, are not yet fully specified by the protocol. Future governance changes to these parameters could materially change risk dynamics, affecting both operator incentives and restaker protections.

• Loan Repayment Volatility
  Operators may repay early to redeploy capital to higher-yield opportunities elsewhere, ending restaker yield flows abruptly. While capital is returned, fee predictability is reduced, and frequent turnover can impair reinvestment timing, expose idle capital, and erode restaker APY targets.

• Opaque Whitelisting

Cap’s architecture opens high-yield opportunities for restakers but also concentrates non-trivial credit, operational, and structural risks that diverge from traditional AVS profiles. Sustaining performance in this model demands institutional-grade due diligence, continuous real-time monitoring, adaptive hurdle rate governance, and liquidity/risk modeling frameworks that integrate both credit and DeFi-specific tail risks.

To accurately assess Cap’s risk profile and calibrate appropriate security thresholds, we parse the protocol’s design into key execution, security, and reputation vectors, then apply a credit-based target stake model.

→ Evaluates whether the protocol reliably enforces intended outcomes via its operational layer — in Cap’s case, whether liquidation logic, credit terms, and restaker protections are mechanistically executed through DeFi primitives.

Cap’s model replaces validator consensus with objective enforcement of bilateral credit agreements. Operator behavior is bounded by fixed KPIs—timely repayment and LTV maintenance—enforced through automated liquidation triggers and deterministic slashing. Execution integrity depends on precise parameterization and the reliability of price oracles and health factor calculations; failures here propagate directly to reserve solvency and restaker losses.

→ Assesses how state agreement and enforcement guarantees are established. For AVSs without internal consensus, this captures the sufficiency of external networks and deterministic logic in replacing native fault-tolerant coordination.

Cap has no internal consensus. State enforcement is handled by deterministic slashing and liquidation logic, with liveness and security inherited from its securing networks: EigenLayer, Symbiotic, Ethereum, and MegaETH. This shifts execution risk outward: disruptions or failures in these networks could stall or block enforcement during operator defaults.

→ Measures how clearly slashing is defined, how deterministically it is triggered, and whether it protects restakers proportionally.

Cap defines slashing deterministically: repayment failure or a breached collateral NAV health factor automatically triggers penalties. Attribution is objective, and slashing may be absolute, as the delegated amount can be fully liquidated, which raises fault severity and incentivizes conservative vault design. The structure aligns Cap with Type III “self-enforcing” stablecoins, where enforcement and yield-generation are embedded in protocol logic rather than dependent on governance discretion.

→ Captures the security audit status, transparency, and structural complexity of the codebase.

Cap has announced five completed audits, and its codebase has been made publicly available. The architecture falls in the medium-to-high complexity range: a moderate number of contracts with clear separation of concerns (vault, reserve, delegation, rate models, liquidation). The interactions between modules are non-trivial due to on-chain/off-chain dependencies, auction mechanics, and dynamic rate logic, yet the structure remains cohesive and avoids deep inheritance or excessive dependency chains. Integration points with Symbiotic/EigenLayer and external yield sources add operational complexity.

→ Surfaces long-term execution risk by gauging operational history, governance quality, and the founding team’s track record.

Even though it is early-stage, Cap is led by reputed builders with prior DeFi and stablecoin experience at Beefy Finance, Frax, and QiDAO. While the architecture is sound in principle and reputation benefits from the team’s track record, operational resilience and liquidation mechanics remain untested under scale or stress.

In security terms, Cap’s design also distances from traditional AVSs: operators do not commit bonded stake to a validator set, but instead secure restaker delegations and borrow capital against which they generate yield. The new paradigm shifts our original Target Stake concept from a validator-level stake threshold to a DeFi-native security bound. The delegated loan principal can be liquidated and slashed in full at any time, instantly collapsing the operator’s open positions and erasing any meaningful yield retention.

Cap is a natively exogenous AVS, in that operators operate purely within DeFi by earning yield entirely outside the protocol through strategies, and security depends solely on repayment of borrowed stablecoins.

In this structure, PfC is not generated by endogenous manipulation (e.g. MEV, token minting through bug exploits, consensus faults) but by exogenous credit risk—the potential misappropriation or loss of delegated collateral via corruption.

For an operator $O$:

Where:

• $\text{PfC}_O$: Maximum profit attainable from corrupting

• $D_O(t)$: Delegated collateral at time $t$

The $\text{PfC}_O$ metric is bounded by the maximum delegation an operator controls at any point in time. In Cap, liquidation to recover the principal unwinds the operator’s positions, so the generated yield evaporates, preventing undue seizure.

The CoC is the total economic position forfeited upon default:

Where:

• $D_O(t)$: Delegated collateral at time $t$

• $Q$: Qualitative deterrents — credit-risk profile of restakers, reputational cost, legal liabilities, and opportunity cost (lost future income)

Both PfC and CoC share $D_O(t)$, which is sufficient to deter an attack; the additional deterrents $Q$ represents only reinforce the gap, making $CoC > PfC$ unquestionable.

Anchoring $Q$ in Credit-Risk Data

Institutional borrowers maintain records with default probabilities and credit spreads from sources such as FRED and S&P. Cap uses these to set baseline delegation pricing, aligning operator borrowing costs with market underwriting standards. Restaker risk appetite directly shapes $Q$ by adjusting delegation size, fee demands, and exposure preferences: Cautious restakers may limit allocations, demand higher fees from lower-rated operators, or avoid correlated exposures entirely—reducing the size of profitable defaults while raising the cost of capital for riskier borrowers.

Monitoring how delegations diverge from credit-spread benchmarks provides insight into delegation health. Over-allocation to CCC-rated operators, concentration in a single tier, or preference for low return-per-unit-of-risk exposures all signal shifts in discipline.

Cap’s baseline framework for restaker allocations, derived from credit spreads and default probabilities:

Learn more here.

Note: We strongly advise conducting dedicated credit-risk evaluations from specialists to fully inform underwriting decisions and ensure thorough due diligence before committing any stake to Cap.

These considerations allows $Q$ to capture not only static deterrents like reputational or legal liabilities, but also the dynamic enforcement applied by restakers in response to fast-pacing credit conditions.

Credit-based cryptoeconomic security against intentional default requires:

Cap’s system ensures this inequality holds. Cryptoeconomic security is anchored at the DeFi credit-allocation layer rather than the validator-bonding restaking layer. Such distinction, coupled with the protocol’s ability to liquidate the entire delegated loan principal instantly, collapses any open positions and erases potential profit extraction. The result is a design that is inherently resistant to corruption-driven attacks by eliminating the feasibility of retaining gains from malicious behavior, and not by increasing “Target Stake” requirements.

This section maps Cap’s integration points across the DeFi and restaking stacks: SSPs, Restakers/LRTs, and Operators, focusing on capital flows, incentive design, and risk propagation. Cap’s robust use of off-chain legal enforcements, deterministic on-chain slashing, and delegated-collateral underwriting alters conventional assumptions about AVS risk delegation and alignment.

Cap leverages EigenLayer and Symbiotic not only for capital provisioning but as slashing enforcement backbones. SSPs route restaker delegations, execute deterministic on-chain slashing on Ethereum upon a Cap operator default, and, in a slashing event, redistribute proceeds to Cap’s reserve and restakers, preserving cUSD peg integrity. Prior to redistribution, Cap’s own contracts execute in-protocol Dutch auctions of liquidated collateral.

SSP selection by Cap should weigh collateral health, slashing determinism, latency, and observability, alongside withdrawal-epoch flexibility that supports restaker liquidity needs. SSP integrations should prioritize modularity, so that stake sourcing, governance exposure, and slashing semantics remain separated and customizable by Cap.

By integrating Cap, SSPs convert idle restaked assets into credit-backed collateral, unlocking yield potential and improving stake productivity while increasing TVL stickiness. The trade-off being duration-locked delegations (loan term plus any SSP withdrawal latency) which can add liquidity stress.

SSPs must monitor delegation concentration, withdrawal queue congestion, and slashing throughput to ensure that Cap-linked collateral sourcing does not impair network-level liquidity guarantees. Importantly, restaked capital supporting Cap loans is never idle on aggregate: when loan proceeds are held in reserve rather than deployed in DeFi strategies, Cap can route associated stablecoin liquidity into fractional reserve strategies, enabling yield capture while preserving redeployment optionality for restakers.

Cap’s whitelisting process should prioritize operator reputation, robust financial records (including PnL transparency), and proven on-chain liquidity management. Slashing is triggered by repayment failure or collateral value breaches below safety thresholds, making collateral volatility a first-order risk parameter.

Cap should calibrate hurdle rates conservatively, track loan repayment volatility, and monitor insolvency indicators such as NAV-to-loan deterioration, liquidity coverage shortfalls, and correlated drawdowns across operator strategies. Delegation caps and a diversified operator base help curb centralization and systemic risk.

With Cap, Operators access a zero-cost-basis credit line, effectively amplifying leverage beyond their own equity and retaining any surplus yield after hurdle rate and restaker fees. This structure can significantly enhance returns, but also magnifies tail risk when strategy margins are thin or collateral prices volatile. Borrowing capacity hinges on active restaker delegations, with Cap applying a grace period for replacing withdrawn delegations mid-loan to avoid forced unwinds.

Operators must meet hurdle rates and restaker fees, actively manage mark-to-market exposures, and maintain liquidity buffers. While surplus yield after obligations can incentivize performance, poorly set hurdles risk driving excessive leverage. Operators should internalize Cap’s systemic risk profile as an AVS and protocol (including slashing propagation scenarios) and avoid clustering in the same high-yield strategies as their peers.

Cap must make onboarding seamless yet protective, providing comprehensive legal templates that embed restaker safeguards, collateral rights, operator recourse provisions, and redemption and service-level guarantees. These agreements should be comprehensive and extensible, giving restakers a standardized but customizable foundation for risk control. Cap should also actively advise restakers on monitoring requirements, including concrete metrics for operator health, collateral liquidity, and slashing exposure, so that oversight is continuous, not reactive.

Cap must remain conscious of restaker concentration levels and communicate the slashing covariance risk inherent in correlated DeFi strategies (covered above in Covariance Risk in DeFi Strategies), reinforcing the need for operator diversification and mitigation of principal–agent misalignments. Restakers should be explicitly warned of capital lock-in through the combined loan term and withdrawal latency, as well as the full-loan slashing risk, including liquidations triggered by declines in restaked collateral value. A severe credit event can wipe principal entirely, making it imperative for LRTs to underwrite operator performance rigorously, maintain liquidity buffers, and size Cap allocations conservatively to protect vault stability.

Attracting and retaining a diversified base of restakers is the core engine of Cap’s recurring-yield flywheel and the foundation for sustaining durable, scalable returns.

Restakers and LRTs, including Swell, Ether.fi, and Renzo, gain diversified yield streams and exposure to non-infra AVS risk, improving portfolio balance. However, potential full-loan slashing means they must independently monitor Cap’s redemption reserves (that honor withdrawals), price-volatility exposure, cUSD peg integrity, and assess Cap’s choices on operator whitelisting, oracles, and bridges—especially in cross-chain contexts.

Position sizing should be counseled and stress-tested against adverse credit and market scenarios to safeguard vault or operator solvency. Even when not deployed into active DeFi strategies, capital custodied in Cap’s reserves can be routed into fractional reserve lending, allowing idle stablecoins to earn yield while remaining immediately available for withdrawal or DeFi reallocation—an added layer of capital productivity.

Delegated capital in Cap functions as underwritten credit, not passive stake. Any principal–agent issue is essentially mitigated through robust bilateral legal agreements that formalize loan terms, recourse provisions, borrowing limits, and redemption rights. Restakers/LRTs should conduct thorough vetting and diversify operator exposure to reduce correlated loss potential.

Operators must recognize that defaults, underperformance, or slashing events can impact multiple restakers and LRTs simultaneously. Strategy concentration risk escalates when operators converge on similar high-yield trades, creating dangerous covariance clusters. To mitigate, operators should diversify across instruments, maturities, and markets to limit both single-point and correlated failure modes.

Succinctly, the durability of the system depends on disciplined underwriting, vigilant monitoring, and deliberate diversification at every link in the chain. When one participant underestimates these dynamics, the impact can cascade across the entire stack.

Cap shifts the foundation of AVS design from on-chain validation to securing credit agreements. Its model blends on-chain liquidation rules with market-based credit pricing, binding restakers, institutional operators, and stablecoin users in a single system of incentives and enforcement. The appeal is straightforward: high capital efficiency and stable yields delivered through programmable, verifiable guarantees. The risk is equally objective: exposure to credit defaults, liquidity stresses, and operator underperformance.

The system’s durability rests on cUSD demand, credit discipline, and delegation diversification. Restakers must calibrate delegation to operator creditworthiness, operators must manage strategies within risk thresholds, and the protocol must enforce parameters that ensure demand and prevent concentrated exposures.

Cap’s robust, innovative design as a restaking-backed credit protocol positions it as a strong candidate for a yield-bearing stablecoin system with scalable adoption potential.

Check Cap's website and Twitter.

Follow us on X and subscribe!