Adversarial Due Diligence Report: cap.app (Cap Protocol / cUSD)

Report Date: 2026-04-04 Analyst: DeFi Adversarial Research Agent Framework: Guilty-Until-Proven-Innocent Confidence Level: Medium-High Overall Risk Verdict: MEDIUM RISK (with systemic architecture caveats warranting MEDIUM-HIGH)

⚠️ LEAD FINDING — THE “BLOW-UP-PROOF” GUARANTEE IS AN OFF-CHAIN LEGAL CLAIM BACKED BY CONFLICTED PARTIES

Cap’s central marketing claim is that cUSD is “blow-up-proof” because operator defaults are covered by restaker slashing and, beyond that, by off-chain legal agreements with regulated financial institutions. Independent risk firm Chaos Labs (assessing the protocol for Renzo Protocol integrations) concluded that “recovery depends heavily on off-chain legal agreements rather than on-chain protections” and that “a credit event could result in losses equaling the full loan value.”

The most structurally alarming element: the “regulated financial institutions” who serve as operators — Franklin Templeton, Susquehanna, Flow Traders, IMC Trading, GSR, Nomura’s Laser Digital — are simultaneously equity investors in Cap Labs. These firms collected equity upside from Cap’s growth, AND they are the primary borrowers of user funds from the protocol. If an operator-investor fails, users suffer the primary loss while the entity in default is the same entity that helped capitalize and promote the protocol. This is a textbook conflict of interest that the protocol does not address publicly.

Compounding this: over 80% of all cUSD reserves (~$360M) are currently deployed into Aave V3 — a single third-party DeFi protocol. The protocol’s “safe” idle reserve layer is itself a concentrated DeFi counterparty risk. An Aave exploit or freeze would be existential for cUSD backing.

Executive Summary

CategoryFinding
Protocol TypeCovered-credit yield stablecoin (cUSD / stcUSD)
ChainEthereum mainnet; MegaETH (L2, launched Feb 2026)
TVL~$271M (DeFiLlama); self-reported $500M at peak Jan 2026
Mainnet LaunchAugust 19, 2025
FounderBenjamin Sarquis Peillard (@Benjamin918_); named, verifiable
Prior RoleCOO at Hashing Systems (crypto infrastructure)
Funding$11M total (Pre-seed $1.9M + Seed $8M + Community $1.1M)
Lead InvestorsFranklin Templeton, Triton Capital, Susquehanna, Flow Traders, Nomura/Laser Digital, GSR, IMC Trading, RockawayX
Audits8 engagements: Zellic, Trail of Bits, Electisec, Spearbit (×2), Recon, Sherlock, Certora
CAP TokenNot yet fully launched; 40% insider allocation (team 20% + investors 20%); no vesting schedule disclosed
Overall RiskMEDIUM (no exploits; systemic design risks and conflicts unresolved)

Team Assessment

Founder: Benjamin Sarquis Peillard

  • Identity: Fully named and public. Twitter: @Benjamin918_, LinkedIn active, featured on Epicenter podcast (Episode 618), multiple media profiles (Fortune, CoinDesk, The Block quotes).
  • Claimed history: “3x crypto founder, OG DeFi contributor” per LinkedIn post.
  • Verifiable prior role: COO at Hashing Systems (crypto infrastructure company). F6S profile confirms this title.
  • Unverifiable claims: The “3x founder” claim cannot be independently verified from public records. No names or outcomes of the prior two ventures were found in open sources. Hashing Systems itself is not a well-documented public company — no website, no CrunchBase listing found.
  • Argentine background: @gsarquis (Gonzalo Sarquis) is active on Twitter and appears to be a relative — consistent with a Latin American background for Benjamin Sarquis Peillard. Not a red flag, merely context.
  • Public accountability: High. Founder has given podcast interviews, appeared in mainstream financial press (Fortune, Fortune Crypto), and is named in regulatory filings context. This is a meaningful legitimacy anchor.

Team Transparency

  • GitHub org: github.com/cap-labs-dev — 7 public repositories, 0 public members listed, membership is private.
  • Publicly named team members: Only the founder is named in press coverage. No CTO, Head of Engineering, or other named team members appear in search results.
  • Known partnerships: WisdomTree (institutional asset manager — Cap provides protected yield to WisdomTree users), Aave (reserve deployment), RedStone (oracle provider), EigenLayer and Symbiotic (restaking infrastructure).
  • Assessment: Single named public face is a moderate concern for a $271M+ protocol. The underlying technical team is anonymous.

Third-Party Consensus

Independent Risk Assessment — Chaos Labs (Renzo Protocol Governance)

Chaos Labs conducted an independent risk briefing specifically evaluating Cap protocol’s risk for the Renzo ecosystem. Their documented concerns:

  1. “Full loan value” slashing exposure: A credit event can produce losses equal to the entire borrowed amount — far exceeding the partial slashing common in PoS slashing conditions.
  2. Active position management required: Restakers must continuously monitor health factors and inject collateral during stress events — a demanding operational requirement.
  3. Off-chain legal agreements as primary protection: On-chain mechanisms cover collateral up to the staked amount. Beyond that, recourse is contractual, not algorithmic.
  4. Liquidity lockup: Delegated stake is locked for the full loan duration; withdrawals require loan completion plus additional waiting periods.
  5. Undefined governance for slashing parameters: Liquidation bonuses, grace periods, and key parameters lack defined governance procedures — can change without restaker input.

Serenity Research (Independent Research Substack, August 2025)

Described Cap’s structure as analogous to a Credit Default Swap (CDS) on-chain — not a CDO, but a CDS. Restakers sign legal agreements to cover defaults and post on-chain collateral. The structure works as long as (a) restakers remain solvent and (b) legal agreements are enforceable across jurisdictions. Neither is guaranteed.

OAK Research (MegaETH Protocol Report)

Published analysis confirming the three-sided marketplace model (minters, operators, delegators) with yield derived from idle reserves deployed to Aave and Fluid. Acknowledged the EigenLayer dependency as a systemic risk factor.

Positive Coverage

  • The Block, Fortune, CoinDesk: Standard funding announcement coverage; uniformly positive tone.
  • Blocmates Guide (August 2025): Detailed explainer; notes the model is genuinely novel but acknowledges restaker risk.
  • Aave Blog: Confirms the Cap cUSD integration on Aave; treated as a positive signal by Aave governance.

On-Chain Findings

Protocol Architecture

Cap operates as a three-sided credit marketplace:

  • Minters: Deposit USDC/USDT (1:1) → receive cUSD; stake cUSD → receive stcUSD (yield-bearing)
  • Operators: Whitelisted institutional borrowers (Franklin Templeton, Susquehanna, Flow Traders, etc.) — borrow cUSD from the vault, generate yield off-platform
  • Delegators/Restakers: Post overcollateralized restaked assets (ETH via EigenLayer or Symbiotic) → vouch for operators; slashed if operator health factor < 1

Reserve Deployment — Aave Concentration Risk

This is an active, observable, high-severity risk:

  • Over 80% of cUSD reserves (reported: $300M–$360M) are deployed into Aave V3 Core Ethereum market.
  • Cap’s “Fractional Reserve” contracts auto-deploy idle capital to Aave to earn base yield for stcUSD holders.
  • This means a critical path to cUSD depeg: Aave V3 exploit → reserve loss → cUSD undercollateralization.
  • Cap acknowledges Aave as a “protocol risk” in their documentation but frames it as low-probability.
  • Verdict: Concentration of 80%+ reserves in a single protocol contradicts the stated “blow-up-proof” design principle. This is the most immediately verifiable systemic risk.

Contracts and Upgradeability

  • GitHub: github.com/cap-labs-dev/cap-contracts — 16 stars, 8 forks, updated April 4, 2026 (today).
  • Repository includes check-proxy-implem.txt and check-access.txt files — indicating upgradeability/access control was considered during development.
  • The precise proxy configuration (UUPS vs. TransparentUpgradeableProxy, timelock duration, admin addresses) could not be confirmed from public search results. Cap’s audit reports in cap-audits suggest the contracts are within Sherlock’s competitive audit scope.
  • Sherlock audit (July 2025): 0 open findings. Scope: Symbiotic integration; Mint, Burn, Redeem, Liquidate, Repay, RealizeInterest. Accepted known issues: 1-wei rounding in withdrawals (seeded), debt token index rounding (+1 wei to total supply).

Oracle Dependency

  • RedStone oracle integration (confirmed December 2025 blog post).
  • RedStone is an established oracle provider but represents an additional off-chain dependency.
  • Oracle failure scenarios: incorrect price feeds could allow undercollateralized borrowing or prevent liquidations.

TVL and Growth

  • DeFiLlama: ~$271M current TVL (all Ethereum).
  • Self-reported (January 2026): $500M peak TVL.
  • Timeline: Mainnet launched August 19, 2025 → $500M in ~5 months.
  • Rapid TVL accumulation without incident is NOT the same as proven security. The protocol has not been stress-tested through a major market downturn.

cUSD Stablecoin Mechanics

  • 1:1 backed by USDC/USDT at issuance.
  • Reserve assets include: USDC, USDT, tokenized money market funds (MMFs), Aave-deposited USDC.
  • cUSD is a LayerZero OFT asset — bridgeable across chains.
  • Bridge risk: As acknowledged in Cap’s own risk docs, cross-chain use of cUSD introduces third-party bridge and oracle risk.

CAP Token (Governance)

  • Status: Not fully launched as of report date. Public sale registration opened February 5, 2026 via Uniswap CCA mechanism.
  • Token allocation:
    • Ecosystem development: 46.72%
    • Team: up to 20%
    • Investors: up to 20%
    • Community ICO: 10%
    • Echo community sale: 3.28%
  • Vesting schedule: NOT PUBLICLY DISCLOSED for the 40% insider bucket.
  • Insider allocation of 40% with no disclosed vesting is a major governance red flag — if the token launches and unlocks immediately, selling pressure could be severe.
  • “Stabledrop”: $12M in cUSD distributed to early Frontier phase participants (February 2026). Framed as equivalent to ~5% of a $250M FDV. Source of the $12M: Cap’s own protocol treasury (funded by the $11M raise + early yield earnings). This is a legitimate community distribution mechanism, not a Ponzi dynamic.

Red Flags Register

#SeverityFindingEvidence
1🔴 HIGHInvestor-operator conflict of interest — Franklin Templeton, Susquehanna, Flow Traders, IMC, GSR, Nomura are simultaneously equity investors and protocol borrowersFortune, The Block funding coverage; Cap blog posts confirming same entities as operators
2🔴 HIGH80%+ of cUSD reserves concentrated in Aave V3 — single DeFi counterparty failure would be existential for cUSD backingStabledash article; Aave blog; stcUSD mechanics docs
3🔴 HIGH”Blow-up-proof” guarantee relies on off-chain legal recourse — Chaos Labs confirmed primary protection is legal contracts, not on-chain code; credit event can equal full loan value lossChaos Labs Cap Risk Briefing (gov.renzoprotocol.com)
4🟠 MEDIUMTrail of Bits audit NOT in their public publications GitHub — claimed audit at github.com/trailofbits/publications not listed; cannot be independently cross-referenced via firm’s standard publication indexDirect GitHub check of trailofbits/publications
5🟠 MEDIUM40% insider token allocation with no disclosed vesting schedule — if team/investor tokens unlock at or near TGE, severe sell pressure possibleCap docs token page; airdrop review sites
6🟠 MEDIUMUndefined governance for slashing parameters — liquidation bonuses, grace periods changeable without restaker input; restakers locked in for full loan durationChaos Labs Risk Briefing
7🟠 MEDIUMMegaETH deployment on 2-month-old mainnet — cUSD launched on MegaETH (mainnet: Feb 9, 2026) introducing compounded L2 infrastructure risk on brand-new chainCoinDesk MegaETH mainnet coverage
8🟠 MEDIUMOnly founder is named publicly; technical team is anonymous — GitHub org has 0 public members; no CTO, Head of Smart Contracts, or other key personnel namedGitHub cap-labs-dev; media coverage
9🟠 MEDIUM”3x founder” claim unverifiable — Benjamin Sarquis Peillard’s prior two ventures cannot be identified or assessed from public recordsF6S profile; no corroborating sources
10🟡 LOWHashing Systems background unclear — “COO at Hashing Systems” as prior role appears on F6S but Hashing Systems has no public web presence, CrunchBase listing, or searchable historyF6S member profile
11🟡 LOW$500M TVL growth in 5 months untested through market downturn — protocol has never operated through a significant market stress event or liquidity crisisDeFiLlama; self-reported cap figures
12🟡 LOWLayerZero OFT bridge risk — cUSD bridgeable across chains; cross-chain scenarios introduce third-party bridge failure modesCap docs risk page
13🟡 LOWRedStone oracle dependency — single oracle provider; incorrect feeds could allow undercollateralized loans or blocking of legitimate liquidationsRedStone blog post Dec 2025
14🟡 LOWCertora audit not in their public SecurityReports GitHub — audit report exists as CDN PDF but is not indexed in github.com/Certora/SecurityReports as of review dateDirect GitHub search
15🟡 LOWMegaETH is not formally EVM-equivalent — ultra-high-throughput L2 may have execution differences; new audit surface for contracts deployed thereMegaETH documentation

Unresolved Questions

  1. Exact collateralization ratio: What is the minimum required overcollateralization ratio for operator loans? Is it hardcoded or governance-adjustable? What happens if restaker collateral value drops faster than the grace period allows liquidation?

  2. Off-chain legal agreements — jurisdiction and enforceability: Which jurisdictions govern the “Guarantor Agreements” with operators? If a Flow Traders entity in Amsterdam defaults, how long does legal recourse take? Who bears the loss in the interim?

  3. Operator whitelist process: Who controls the whitelist? How are operators added or removed? Is there a timelock? Can the Cap Labs team add themselves or affiliated entities as operators?

  4. Vesting schedule for team/investor tokens: When do the 40% insider CAP tokens unlock? Are there cliffs? This is the single most important undisclosed tokenomics parameter.

  5. Hashing Systems: What was this company? When did it operate? What was Benjamin’s role/tenure? Any prior project failures?

  6. Trail of Bits audit content: Why is the May 2025 Trail of Bits audit not listed in their public publications index? What were the findings? Were all issues resolved?

  7. Proxy/upgradeability config: What are the exact upgrade mechanisms in the deployed contracts? Who controls the admin keys? Is there a timelock and what is its duration?

  8. Aave concentration — is there a cap? Is there a governance-set maximum percentage of reserves that can be deposited in any single protocol? Currently 80%+ suggests no enforced diversification requirement.

  9. Flow Traders as publicly traded operator: Flow Traders (AEX: FLOW) disclosed in a Cap blog post that it uses Cap via EigenLayer. Does this appear in their public financial disclosures? What exposure size is disclosed?

  10. MegaETH contract audit: Were the contracts deployed on MegaETH separately audited, or were only the Ethereum contracts reviewed?

Positive Signals

  • Named, media-facing, publicly accountable founder who has given multiple interviews
  • Franklin Templeton as strategic investor — one of the most compliance-conscious asset managers in TradFi
  • 8 audits from reputable firms over a 9-month period; Sherlock competitive audit found 0 valid issues
  • $271M TVL demonstrates genuine user adoption
  • WisdomTree partnership demonstrates institutional demand for the product
  • Protocol went mainnet August 2025 and has had no exploits in its first 8 months
  • The “Stabledrop” mechanism (distributing $12M of yield as stablecoin, not a volatile token) is a sophisticated community distribution that avoids the pump-and-dump token airdrop dynamic
  • Certora formal verification performed on EigenLayer integration
  • GitHub public and actively maintained

On-Chain Contract Addresses (Partial — From Available Sources)

AssetAddressChain
cUSDNot confirmed in public search resultsEthereum
stcUSDNot confirmed in public search resultsEthereum
GitHubgithub.com/cap-labs-dev/cap-contracts—
Twitter@capmoney_—
Docsdocs.cap.app—
DeFiLlamadefillama.com/protocol/cap—

Summary Verdict

MEDIUM RISK | Confidence: Medium-High

Cap is a structurally ambitious, well-funded, and genuinely novel protocol — but its “blow-up-proof” marketing claim obscures a model that depends on: (a) off-chain legal agreements with counterparties who have conflicts of interest; (b) 80%+ reserve concentration in a single external DeFi protocol (Aave); and (c) restaker collateral being sufficient to cover full operator defaults in a stress scenario.

The protocol has not been tested through a bear market or liquidity crisis. TVL of $271M–$500M in <8 months with no exploits is a positive signal but not a security guarantee. The Franklin Templeton brand provides a legitimacy anchor but does not resolve the structural conflict of interest inherent in investor-operators.

The critical unknown is the vesting schedule for 40% of CAP tokens. If unlocks are near-term and uncliffed, token launch could be an exit event for early insiders.

For users: cUSD itself may be relatively safe for short-term holding given its USDC backing and Aave yield layer. The greater risk is a black swan scenario (Aave exploit + simultaneous operator default + restaker liquidation cascade), which while unlikely, would be existential. The “blow-up-proof” guarantee should be read as “blow-up-resistant with off-chain backstop” — a meaningfully different claim.