Dolomite

Protocol identification — Dolomite

• Name / website / docs: Dolomite; https://dolomite.io; https://docs.dolomite.io.
• Category: DeFi margin-trading, lending/borrowing, and spot-DEX infrastructure.
• Launch date: October 13, 2022, initially on Arbitrum One.
• Chains in scope: Arbitrum One, Berachain, and Ethereum Mainnet. The documented core deployment is also present on other networks, but those are outside this request.
• Native token: DOLO, ERC-20; documented address 0x0F81001eF0A83ecCE5ccebf63EB302c70a39a654 on all supported chains. veDOLO and oDOLO are separate ecosystem contracts. Main contract addresses
• Arbitrum: DolomiteMargin — 0x6Bd780E7fDf01D77e4d475c821f1e7AE05409072.
• Berachain: DolomiteMargin — 0x003Ca23Fd5F0ca87D01F6eC6CD14A8AE60c2b97D.
• Ethereum: DolomiteMargin — 0x003Ca23Fd5F0ca87D01F6eC6CD14A8AE60c2b97D. Verification limitation: The requested two-source cross-check including Dune cannot be completed because Dune MCP is unavailable in this run. Therefore, the address deployments, explorer verification status, and whether the same bytecode is live on each chain are Not verifiable as of October 13, 2026. Explorer pages were also inaccessible to this web fetcher (HTTP 403). The addresses above are documentation-sourced, not on-chain-verified. Fork lineage
• Yes: Dolomite’s core repository is explicitly forked from dydxprotocol/solo; the code retains dYdX/“Solo Margin” lineage.
• Material changes: compiler upgrade; gas-efficient market/account tracking; additional risk controls (oracleSentinel, market caps, interest-rate caps); account-risk overrides/e-mode; and liquidation/operator changes.
• Auditing: Original Solo code was audited by OpenZeppelin and Bramah; Dolomite changes received SECBIT and later Cyfrin review. Modules were separately audited by Zokyo, Guardian, and Cyfrin, but audit scope is version- and module-specific—not blanket assurance.
• Malicious-modification history in similar forks: Not verifiable as of October 13, 2026. A 2022 dYdX npm-package compromise affected software distribution, not the smart contracts; it is not evidence of a malicious Dolomite fork modification.

Dolomite appears to be a real, functioning product portal rather than a static landing page: the live app exposes balances and an Earn page, and the docs describe active deposit/withdraw flows through a DepositWithdrawalRouter and wallet confirmation flow. The documentation also includes developer material and a published subgraph, which is a maturity signal for integrations and user support. There is evidence of product UX, not just marketing copy: docs link directly to the Balances page, explain withdrawal steps, and note that users can deposit, withdraw, and view transfer history in-app.

The docs also reference a redesigned web app and indicate that app guides are still being updated, which suggests an evolving interface rather than a frozen template site. Open API: a full public API was not directly verified from the gathered sources, but the docs do expose developer-oriented contract addresses, a subgraph, and integration guidance; that is stronger than a pure consumer portal, yet it is not the same as a clearly documented open REST API. Not verifiable as of 2026-10-13.

No broken links, fake metrics, or template-site signs were directly confirmed from the available results. Not verifiable as of 2026-10-13.

Dolomite appears to have an active bug bounty program. The program is public in Dolomite’s documentation, asks reporters to email security@dolomite.io, covers any vulnerability not previously disclosed by Dolomite or its auditors, requires confidential handling for 30 days, promises an initial confirmation within 72 hours, and says rewards are determined using the OWASP risk assessment methodology. Public third-party coverage also describes the program as active and says it is run with Immunefi, but that specific platform connection is not confirmed in Dolomite’s own docs.

The exact program start date and any verified bounty-payout results are not verifiable as of 2026-10-13 from the available sources.

Assessment as of October 13, 2026: dependency_failure_active: false — no active Dolomite, oracle, bridge, or integrated-protocol failure was identified in the reviewed sources. This is not an on-chain attestation. Dolomite’s principal external dependencies are asset issuers and integrated DeFi protocols, not custodians.

Documented Arbitrum integrations include GMX/GLP/GM assets, Pendle PTs, Jones, Plutus, Abracadabra and other vault or LP wrappers. Failure of an underlying protocol, vault, liquidation mechanism, or reward system could impair collateral value or make withdrawals/liquidations unavailable. Oracle risk: ordinary assets generally use Chainlink; complex assets may combine Chainlink with protocol-specific calculations or Pendle TWAPs. This creates dependence on Chainlink availability/configuration, TWAP liquidity, GMX accounting, and adapter correctness.

A stale, manipulated, or economically incorrect oracle could cause under-collateralized borrowing, bad debt, or unfair liquidations. Dolomite mitigates this with isolation mode, collateral-only mode, supply caps and restricted debt baskets, but these controls do not eliminate oracle or adapter risk. Stablecoin/LST/restaking exposure: documented markets include USDC, USDT, HONEY, rUSD, sUSDe and other Berachain assets; Arbitrum markets include PT-wstETH, PT-eETH, PT-rsETH, uniBTC and GMX-related assets. Risks include issuer insolvency, depeg, redemption freeze, validator/slashing events, restaking failure and liquidity mismatch.

HONEY additionally depends on Berachain collateral-vault governance and collateral quality. Bridges and custody: user-facing documentation references Arbitrum’s canonical bridge plus Across, Hop and Synapse. Protocol-level bridge balances, bridge concentration, custodians, CEX/MM exposure, and RWA issuer/SPV exposure are not verifiable as of October 13, 2026. Dolomite’s virtual-balance model and proxy vault integrations add smart-contract and accounting dependencies, but do not establish institutional custody.

A retired Ethereum-era Dolomite contract was exploited on March 20, 2024; reporting states the current Arbitrum deployment was unaffected. This remains relevant legacy-contract and approval risk, not evidence of an active failure. max_exposure_pct: null — on-chain exposure cannot be verified because Dune MCP was unavailable; no percentage is inferred.

Dolomite appears to be non-custodial: the protocol terms state that Dolomite.io does not have possession, custody, or control over user crypto, does not store or receive crypto, and users retain control when interacting with the smart contracts. The protocol also describes a smart-contract-based, self-custodied model where transactions are executed peer-to-peer through open-source contracts. The architecture is not fully segregated at the protocol level in the traditional custodial sense: the whitepaper says balances in the protocol are pooled amongst all users, even though control remains with the users’ wallets and the system is governed by smart contracts.

Administrative control is organized through a multi-signature / delayed multi-signature setup for protocol parameters, with governance and operators able to manage listings and risk settings, but this is operational control rather than custody of user assets. For the requested fields: withdrawal_paused is Not verifiable as of 2026-10-13 from the provided sources, and segregated_assets is false if interpreted as separate, individually ring-fenced custodial asset segregation; the protocol instead uses pooled protocol balances under self-custody smart contracts.

Dolomite uses a multisig-plus-timelock administrative model, supplemented by role-based permissions.

• Core admin authority: Dolomite documentation describes a Gnosis Safe controlling a delayed multisignature wallet. The documented launch configuration was a 2-of-3 Safe, with the delayed wallet enforcing a transaction timelock. However, that documentation describes the initial/launch arrangement; the current signer set, threshold, and delay are Not verifiable as of October 13, 2026 without on-chain verification.
• Role separation: The current Modules code defines distinct roles, including DEFAULT_ADMIN_ROLE, SECURITY_COUNCIL_ROLE, LISTING_COMMITTEE_ROLE, EXECUTOR_ROLE, and BYPASS_TIMELOCK_ROLE. Permissions can be restricted by destination contract and function selector, limiting what operational roles may submit or execute.
• Timelock and emergency path: Transactions are recorded on-chain and normally require the configured delay before execution. A separately assigned bypass role can execute eligible transactions immediately; the code prevents an address from simultaneously holding the default-admin and bypass-timelock roles.
• Governance and treasury: veDOLO holders vote on major governance matters, while passed proposals are implemented by Leavitt Innovations. Treasury operations are described as being executed by a treasury multisig. This is an operational/governance layer, not evidence that governance directly controls every protocol admin key.
• Key custody: Public sources do not disclose whether individual signer keys use hardware wallets, MPC, HSMs, geographic separation, signer rotation, or formal recovery procedures. Not verifiable as of October 13, 2026. Risk assessment: The model reduces single-key risk through threshold signing, delayed execution, and scoped roles, but residual governance/key-person risk remains because the current multisig membership, threshold, timelock values, and emergency signers are not independently verifiable here for Arbitrum, Berachain, or Ethereum. Exact chain-specific organization: Not verifiable as of October 13, 2026.

Scope/as-of: October 13, 2026. Dune was unavailable; therefore current on-chain role membership, proxy-admin ownership, decoded admin events, timelock delay, renunciation status, and withdrawal-exit testing are Not verifiable as of October 13, 2026. Key addresses

• DolomiteMargin core: Arbitrum 0x6Bd780E7fDf01D77e4d475c821f1e7AE05409072; Berachain and Ethereum 0x003Ca23Fd5F0ca87D01F6eC6CD14A8AE60c2b97D.
• DolomiteOwnerV2: Arbitrum and Berachain 0xC2B66E247daE5Ee749Ae1d827190115F3653dE06. Ethereum deployment: Not verifiable as of October 13, 2026.
• Berachain core proxy examples: DepositWithdrawalProxy 0xd6a31B6AeA4d26A19bF479b5032D9DDc481187e6, GenericTraderProxyV1 0xb50BcDFC914e0AfB484Dee621F49010862Fb928d, PartiallyDelayedMultiSig 0x52d7BcB650c591f6E8da90f797A1d0Bfd8fD05F9. Architecture / admin risk ``text Users/assets -> immutable DolomiteMargin core -> mutable proxy/module layer -> traders, liquidators, vaults, registries, oracles -> DolomiteOwnerV2 / multisig-controlled administration ` The core is described as immutable, while modules and several proxy contracts are mutable/upgradable. This makes upgradeable = true` for the overall deployment, even if core accounting logic is not upgradeable. Documented privileged functions include market closure/caps, interest-setter and oracle changes, risk parameters, global-operator approval, and excess-token withdrawal. A malicious global operator could put all user funds at risk; this is the principal admin/key-compromise path. Admin-controlled modules or oracle changes could also freeze borrowing, distort solvency, or redirect trades. Users should generally be able to withdraw solvent balances, but this is Not verifiable as of October 13, 2026. The documentation describes an initial 2/3 Safe behind a delayed multisig and a one-day launch delay, with bypass functions for selected emergency/configuration actions; the current delay and actual signers are Not verifiable as of October 13, 2026. Audits exist for Core and Modules, but audit scope differs and does not prove every current deployment is audited.

Bramah Systems — dYdX Solo Assessment. Report delivered February 18, 2019. Scope: all Solidity code at commit 77917ea plus relevant TypeScript tests, focused on trading actions, token theft/freezing, arithmetic and trading interference.

Severity classes were not assigned. The report states identified issues had been addressed and no significant security concerns remained. Bytecode match to deployed Dolomite contracts: no; covers deployed code: false.

OpenZeppelin — Solo Margin Protocol Audit. Published April 30, 2019. Scope: dYdX Solo contracts/protocol folder, commit 17df84d; original upstream code, not a Dolomite deployment.

Findings: 0 critical, 2 high, 2 medium, 10 low/notes. Highs included experimental ABIEncoderV2 and malicious AutoTrader fund theft. Fixes were documented for several findings.

Bytecode match to deployed Dolomite contracts: no; covers deployed code: false.

Cyfrin — Dolomite Proof of Liquidity (POL) Vaults v2.0. Published April 24, 2025. Scope: Berachain POL vaults, isolation-mode vaults, reward registries, Infrared integration, traders, proxies and price oracle.

Findings: 0 critical, 1 high, 1 medium, 3 low, 3 informational and 1 gas optimization. H-1 and M-1 were resolved; L-1 and L-2 acknowledged; L-3 resolved. Bytecode match to deployed contracts was not independently established.

Cyfrin — Dolomite Margin Core Audit. Published August 23, 2023. Scope: latest Core updates at the reviewed commit. Findings: 0 high, 5 medium, 6 low and 4 informational. Final remediation status was not independently verifiable from accessible report metadata. Bytecode match to each deployed contract was not independently established; covers deployed code: not verifiable.

Guardian — GMX V2 Module. Published January 11, 2024. Scope: Dolomite GMX V2 module on Arbitrum, commit 1a91bbd. Findings: 2 critical, 4 high, 13 medium, 15 low. Guardian reports 30 resolved and 4 acknowledged; it also records three additional critical issues found during fix review. Bytecode match to deployed contracts was not independently established.

Zokyo — Dolomite Protocol Modules: GLP & Isolation Mode. Published April 19, 2023. Scope: Dolomite Margin Modules covering GLP and Isolation Mode functionality.

Critical/high/medium finding totals and fix status were not reliably verifiable from accessible primary report text. Bytecode match to deployed contracts was not independently established; covers deployed code: not verifiable.

SECBIT Labs — Dolomite Margin Audit. Published August 2, 2021. Scope: Dolomite Margin Core before Dolomite’s Arbitrum One launch.

Dolomite identifies it as a pre-launch audit. Severity-specific critical/high/medium totals and final remediation status could not be reliably extracted from the published PDF. Bytecode match to deployed contracts: not independently established; covers deployed code: false.

Founders and team. Dolomite is associated with Corey Caplan and Adam Knuckey, who reportedly met at Lehigh University and began working together on decentralized-exchange technology in 2018. Caplan is described as co-founder/CEO or president, with responsibility for smart-contract development and technical direction; Knuckey is described as co-founder/COO, leading frontend and operations. The protocol was developed through Leavitt Innovations, according to multiple industry profiles. Prior projects and outcomes. The founders’ principal documented prior project was an early DEX built on Loopring; sources say they built an Android DEX and later used that experience and infrastructure to develop Dolomite.

This is meaningful technical experience, but the available evidence does not independently establish the commercial outcome of the Loopring venture beyond descriptions such as “moderate success.” Prior hacks or exploitable failures: Not verifiable as of October 13, 2026. Public versus anonymous. The two principal founders are publicly named, and an Arbitrum governance application lists additional contributors using both legal-name and pseudonymous identities, including “Bowtied Oriole” and “Kai Masters.” Public biographies support a real technical and operating team, but they are largely self-reported or reproduced by secondary platforms; independent employment, identity, and track-record verification is limited. Credibility and operating substance. Positive indicators include a multi-year development history, documented Loopring work, Arbitrum governance participation, and deployment across several EVM networks. However, the search evidence does not independently verify current revenue, employee count, audited corporate financials, or the claimed scale of assets. Any protocol-originated or promotional figures should therefore be treated as unverified marketing claims.

On-chain TVL, holder concentration, treasury, and cross-chain exposure are Not verifiable as of October 13, 2026 because on-chain verification was unavailable. Office and jurisdiction. F6S lists Caplan in New York City and gives Dolomite/related company information associated with George Town, Cayman Islands. The Arbitrum governance material identifies a Dolomite Foundation director, but the available sources do not establish a staffed physical office, beneficial ownership, incorporation documents, or whether operations are substantively onshore or offshore. Real office and operating jurisdiction: Not verifiable as of October 13, 2026. Assessment. The evidence is more consistent with a genuine founder-led DeFi engineering project than a purely cosmetic web front, but corporate substance and financial transparency remain insufficiently verified. The principal risk is not founder anonymity; it is the limited independent evidence for legal structure, staffing, business economics, and prior-project outcomes.

Dolomite’s reputation is mixed but not associated in the available evidence with a confirmed rug pull, fraud, insolvency event, sanctions designation, or regulatory enforcement action. However, several important risk questions remain unresolved because on-chain verification was unavailable in this review: Not verifiable as of October 13, 2026.

• Team and backing: Public sources identify Corey Caplan and Adam Knuckey as co-founders, with origins in Loopring-related development and a later Arbitrum deployment. Reported financing includes a $2.5 million 2023 round involving Draper Goren Holm, NGC Ventures, Coinbase Ventures, WWVentures and others, plus a reported $900,000 strategic round. These investor claims are based on secondary databases and media-style summaries, not independently verified capitalization records.
• Audits: Secondary sources report reviews or audits involving firms including OpenZeppelin/Zeppelin Solutions, SECBIT Labs, Cyfrin, Zokyo, Guardian and others. The available material does not establish the complete scope, dates, deployed-code correspondence, or remediation status of every review. Customer-facing incident postmortems and regulated-style transparency reporting are described as limited.
• Sentiment and criticisms: The principal reputational concern is governance and concentration risk linked to Caplan’s reported advisory or technical role at World Liberty Financial (WLFI). Commentary has also alleged that WLFI-related, illiquid collateral represented an unusually large share of deposits and could impair utilization or lender withdrawals. These are allegations and risk commentary, not proof of misconduct or insolvency.
• Fraud, rug, insolvency: No reliable result reviewed confirms a completed rug pull, fraud judgment, insolvency, or loss-of-funds incident involving Dolomite. A full incident and exploit check is Not verifiable as of October 13, 2026.
• Legal, regulatory, sanctions: No reliable result reviewed identified a confirmed enforcement action, court finding, sanctions listing, or formal regulatory designation against Dolomite or its named founders. This check is Not verifiable as of October 13, 2026. The unresolved institutional concerns are asset and borrower concentration, conflicts of interest involving WLFI, liquidity under stress, administrator/key-management powers, audit coverage of deployed contracts, and transparency of incident reporting. Reported TVL, deposits, holders, treasury balances and chain-by-chain exposure are Not verifiable as of October 13, 2026 without raw on-chain validation.

Assessment — Dolomite economic model (as of October 13, 2026) Dolomite is primarily a permissionless money market and margin-prime-brokerage layer. Users deposit supported assets, earn utilization-driven lending interest, or place assets into isolated borrow positions as collateral. Borrowers can loop, hedge, or increase directional exposure through Dolomite’s Zap functions.

Positions are not inherently market-neutral: collateral and debt may be different assets, and GM/GLP, LP, options-vault, and Berachain Proof-of-Liquidity integrations introduce external protocol, oracle, basis, and directional risks. Assets and yield source: Yield is mainly borrower interest allocated to suppliers according to utilization and the protocol’s earnings-rate split; integrated assets may also retain their external yield, such as GMX trading/borrow fees, staking rewards, or Berachain PoL rewards. Dolomite states that the protocol does not take additional fees on GM rewards, apart from integration-specific execution fees. Incentives also exist, including points and DOLO-related rewards; therefore headline APYs may be partly subsidized.

The organic percentage is not independently decomposable. Leverage, lock-ups, withdrawals: No protocol-wide lock-up is documented for ordinary balances. Withdrawals are generally immediate subject to available liquidity; if all liquidity is lent out, withdrawals can fail or require waiting/swapping. Isolation-mode integrations can impose special withdrawal mechanics, including documented 24-hour periods for some assets.

Borrow positions support up to 32 assets, isolated collateral/debt risk, liquidations, and a 5% global liquidation penalty on Arbitrum and Berachain, before market premiums. Fees and revenue: Revenue comes principally from the borrower-interest share retained by the protocol and liquidation penalties. DeFiLlama reports approximately $1.31m fees and $261.9k protocol revenue over 30 days; these are aggregator figures, not Dune-verified. TVL: DeFiLlama reports $352.74m total TVL: Ethereum $323.91m (91.8%), Arbitrum $23.70m (6.7%), and Berachain $5.06m (1.4%). Product-level TVL, Dune TVL, and a reproducible TVL trend comparison are Not verifiable as of October 13, 2026 because Dune MCP is unavailable.

DeFiLlama reports +11% over 30 days and average supply APY of 1.59%; asset-level APY history and volatility are Not verifiable as of October 13, 2026. Risk conclusion: Yield sustainability depends on real borrower demand and utilization, while external integrations add layered smart-contract, oracle, liquidity, and market risks. The reported aggregate active-loan figure slightly exceeds TVL, but this is not a valid user-level leverage measure.

Assessment (as of October 13, 2026): Current liquid-reserve value and liabilities are null. Dune MCP was unavailable, so on-chain balances, asset-level composition, and cross-chain exposure were not independently verified: “Not verifiable as of October 13, 2026.” Disclosed treasury/control addresses

• Treasury/Foundation Safe: 0xa75c21C5BE284122a87A37a76cc6C4DD3E55a1D4, listed on Arbitrum and Berachain as the Dolomite Gnosis Safe; the same address is also the DOLO contract owner on Ethereum.
• Delayed administrative multisig: 0x52d7BcB650c591f6E8da90f797A1d0Bfd8fD05F9, listed on Arbitrum and Berachain. Custody and control: Dolomite documentation describes a 2-of-3 Gnosis Safe controlling a delayed multisig, with administrative transactions subject to a timelock. Governance documentation states that treasury allocations are DAO-governed, while execution is performed by treasury multisig signers and core contributors. Composition/policy evidence: A governance proposal dated August 29, 2025 stated that the treasury held “several million dollars” of stablecoins on Berachain, specifically USDC.e, and proposed borrowing $500,000 of WBTC against it, bridging the WBTC to Arbitrum, and supplying it on Dolomite. This is a historical proposal/disclosure, not evidence of current balances or execution. The DOLO whitepaper assigns 9.65% of supply to the Dolomite Foundation as a treasury reserve for development, grants, and initiatives; current token holdings were not independently verified. Attestations: No independent proof-of-reserves, liabilities attestation, or custody report was located. Not verifiable as of October 13, 2026.

Native token: Dolomite DOLO; ERC-20 address on Arbitrum, Berachain and Ethereum: 0x0F81001eF0A83ecCE5ccebf63EB302c70a39a654. veDOLO and oDOLO are separate Berachain contracts. Supply/valuation: Documented maximum supply is 1.0bn DOLO. CoinGecko reports approximately 513.3m circulating, $13.1m market cap and $25.45m FDV as of the latest available data. > Contradiction: Dolomite documentation states 326.65m DOLO circulating at TGE, while Tokenomics.com reports 0% initial float and different current unlocked figures. These are off-chain/aggregated representations; the authoritative current circulating amount and historical unlock execution are Not verifiable as of October 13, 2026 without Dune/on-chain reconciliation. Utility/governance: DOLO provides liquidity and becomes governance power when locked as veDOLO. veDOLO holders can vote on listings, risk parameters, tokenomics, treasury and incentives; lock duration increases voting weight. veDOLO may receive protocol-fee revenue only if the DAO activates that feature. oDOLO is distributed weekly to liquidity providers and must be paired 1:1 with DOLO to obtain discounted veDOLO. Emissions, fees and burns: Scheduled inflation begins after year four at 3% annually, subject to governance; tokens may fund incentives or be burned. veDOLO early exits include a documented 5% burn fee plus a recoup fee.

No documented automatic buyback; revenue sharing is conditional, not guaranteed. Allocations: Community 50.75%; core team 20.2075%; foundation 10.6511%; investors 15.1846%; service providers 3%; advisors 0.2068%. Team vesting is three years with a one-year cliff; investor vesting includes a one-year cliff and three-year vesting. Unlocks: Announced schedule runs from April 23, 2025 to April 24, 2029, including team, investor, advisor and community releases. Whether announced unlocks actually occurred on-chain: Not verifiable as of October 13, 2026. Admin/control risk: The published contract supports owner-controlled minter permissions and CCIP-admin configuration, plus user-triggered burning; no blacklist or transfer-fee function is shown in the published source.

Current owner/minter addresses and concentration of top holders/insiders: Not verifiable as of October 13, 2026. DEX liquidity/listings: Initial venues are Kodiak (Berachain) and Uniswap (Ethereum). A Uniswap Ethereum DOLO/USDC pool showed roughly $208k liquidity in the latest indexed snapshot; complete multi-chain depth and Arbitrum liquidity are Not verifiable as of October 13, 2026.

A Bitcoin move below $10,000 would likely trigger severe collateral stress across Dolomite on Arbitrum, Berachain, and Ethereum because Dolomite liquidates positions when collateral falls below required thresholds, with liquidations designed to occur before an account becomes undercollateralized. The main protocol-level risk is not Bitcoin directly, but the knock-on effect on any BTC-linked collateral, leveraged loops, and broader crypto market liquidity that can accelerate liquidations and bad debt in stressed markets. Dolomite’s published risk model says liquidation occurs when an account’s collateralization falls below the minimum threshold, and its docs describe liquidation as the mechanism that repays debt by seizing collateral plus a penalty.

In a BTC crash to sub-$10k, positions using BTC or BTC-correlated assets as collateral would be the first to deteriorate; positions with other volatile collateral can also be hit indirectly if the crash drives a broad DeFi de-risking wave, oracle lag, or liquidity withdrawal. For a cross-chain view, Dolomite’s governance material confirms activity across Berachain and Arbitrum, with different rates on each chain, implying chain-specific exposure rather than one unified risk pool. I could not verify the live share of TVL or the exact BTC-linked exposure by chain from the available sources, so the chain-by-chain exposure split is Not verifiable as of 2026-10-13.

The most important stress outcome would be whether liquidations can clear into deep enough markets; if they cannot, the protocol can face residual bad debt. That risk is amplified in thin-liquidity conditions, which is consistent with recent market commentary describing liquidations and sharper crypto drawdowns during BTC stress. Bottom line: a sub-$10k Bitcoin would be a *severe* stress scenario for Dolomite, likely causing widespread liquidations and higher bad-debt risk, but the exact protocol solvency impact is Not verifiable as of 2026-10-13 without on-chain exposure data.

Dolomite’s liquidation logic uses a minimum collateralization threshold of 115% for standard markets on Arbitrum and Berachain, and accounts are liquidated when collateral falls below the required threshold. Under a 20% depeg of the largest collateral asset, the impact is not fully verifiable as of 2026-10-13 because the provided sources do not contain an on-chain position snapshot or chain-by-chain collateral composition for Dolomite on Arbitrum, Berachain, and Ethereum. What can be said from the available evidence is that a 20% collateral price shock would reduce that asset’s value to 80% of its pre-shock value; for a position to remain safely above a 115% collateralization requirement after such a move, it would need to have started with materially more than 115% headroom.

The exact bad-debt or liquidation exposure depends on the share of the largest collateral in each account, the debt asset mix, and whether markets have asset-specific liquidation parameters. Chain-specific exposure is also not verifiable as of 2026-10-13 from the supplied sources. One secondary source indicates Ethereum represented about 83% of reported Dolomite TVL in a snapshot, but that is an aggregator-style estimate rather than raw on-chain verification, and the prompt’s required on-chain check is unavailable here.

The key risk takeaway is that Dolomite’s design can still force liquidations on accounts concentrated in the depegged collateral, especially where borrowers are near the liquidation threshold; however, the protocol-wide loss estimate cannot be stated reliably without the current collateral/debt distribution and market-level risk parameters on each chain.

Stress scenario: largest borrower/strategy counterparty becomes insolvent Exposure identification. The largest counterparty, its debt, collateral, and exposure split across Arbitrum, Berachain, and Ethereum are Not verifiable as of October 13, 2026 because Dune/on-chain verification is unavailable in this run. The analysis therefore applies to the largest leveraged borrower or integrated vault on each chain, rather than naming a wallet. Expected loss path 1. Counterparty default or collateral collapse causes its Dolomite account to fall below the required collateralization threshold.

2. A liquidation operator invokes the liquidation path. Collateral is seized at an oracle-based price plus the applicable liquidation spread; the seized collateral is sold or swapped to repay the owed asset.

Dolomite documents a 5% global liquidation penalty on Arbitrum and Berachain, with market-specific premiums. 3. Expected loss = debt remaining after liquidation proceeds, trading slippage, oracle latency, and accrued interest. The loss is amplified if the collateral is thinly traded, isolated-mode, bridged, or rapidly depegs. Who absorbs it

• First absorber: the defaulting borrower, through loss of collateral and liquidation penalty.
• Second absorber: liquidators, who bear execution, gas, slippage, and inventory risk; their compensation is the liquidation spread/reward.
• Residual bad debt: Dolomite’s excess protocol balances—described in the documentation as an insurance-fund mechanism—can be used through vaporization. The smart-contract implementation uses positive/excess tokens to reduce the insolvent account’s negative balance.
• If excess balances are insufficient, lenders/depositors in the affected market bear the remaining economic shortfall through impaired withdrawals or losses in claim value. An automatic, chain-wide pro-rata socialized-loss mechanism is Not verifiable as of October 13, 2026. Smart-contract impact path Oracle price update → account becomes undercollateralized → authorized/global liquidation operator calls Liquidate → DolomiteMargin.operate transfers collateral and debt → trader/liquidator swaps collateral → residual debt, if collateral reaches zero, enters Vaporize → excess positive balances repay part or all of the bad debt. Risk conclusion: the key tail risk is not borrower default alone, but liquidation failure during a fast market move; the unverified chain-by-chain concentration and available excess-token buffer are the decisive unknowns.

For Dolomite, I do not find evidence of committed fraud by the DAO or owners in the sources provided; the clearest verified event is a 2024 contract exploit affecting discontinued Ethereum-era approvals, which Dolomite says it suspended and then used treasury funds to make affected users whole. The available reporting instead points to a *security incident* and remediation, not an insider-fraud finding: Revoke.cash says the attacker exploited an old contract, recovered most funds through negotiation, and the protocol covered the remaining loss from treasury. Dolomite’s own bug-bounty page and governance materials also describe an active security process and DAO-controlled actions, but they do not evidence fraudulent conduct by the DAO or owners.

On the stress-scenario question specifically, the main residual risk is *governance/owner abuse* rather than confirmed fraud. A governance proposal could theoretically redirect protocol-owned liquidity or treasury assets, and Dolomite’s governance docs show the DAO can authorize collateral-based treasury strategies across Berachain and Arbitrum. However, in the materials reviewed, there is no source alleging that Dolomite’s DAO or owners actually misappropriated funds or ran a fraud scheme.

So the risk assessment is: no verified committed fraud case in the provided evidence; governance centralization / malicious proposal / insider-action risk exists in principle, but it is *not substantiated* here.

Dolomite’s primary yield source can be stressed into a negative 30-day condition when borrower demand and/or utilization drop enough that interest income falls below the baseline expectations embedded in the market. Dolomite documents that lending yield is driven by the protocol’s utilization-based interest model, so a weak 30-day period is fundamentally a lower interest-accrual scenario rather than a separate reward stream. For a stress case, the practical interpretation is:

• Base lending APY compresses or turns near-zero if utilization falls sharply, reducing the protocol’s interest income to depositors.
• Any strategy relying on underlying lending venue yield is exposed to the same compression, because the yield is sourced from borrower-paid interest.
• In a liquidity-stress environment, withdrawals may also become constrained until liquidity returns, which can amplify the effect on realized user returns even if accounting yield remains positive. I could not verify on-chain 30-day yield figures, chain-by-chain exposure, or whether Arbitrum, Berachain, or Ethereum currently show negative realized primary yield as of today. Not verifiable as of 2026-10-13. From a risk lens, the key takeaway is that Dolomite’s primary yield is not structurally protected against a weak borrowing environment; if utilization and fee generation fall for 30 days, the yield source can deteriorate materially and may become effectively negative after fees, slippage, or opportunity cost are included. That last step is an analytical inference, not a directly verified on-chain fact.

Governance is not fully decentralized. Dolomite’s documented model gives veDOLO holders votes over listings, tokenomics, risk parameters, treasury allocations and incentives, using Discord discussion, temp checks and quarterly BeraVote votes; stated launch parameters were ~7-day voting, 5% quorum and 1M veDOLO proposal access. However, operators may independently perform listings and risk updates, and execution is assigned to Leavitt Innovations.

This makes DAO control meaningful for selected policy decisions but not control of all upgrades/admin powers; classify DAO governance as false. Contract control: DolomiteOwnerV2 is deployed on the listed networks and defines DEFAULT_ADMIN, SECURITY_COUNCIL, LISTING_COMMITTEE, EXECUTOR and BYPASS_TIMELOCK roles. Mutable proxy contracts can be modified through Dolomite Margin by its owner.

The current role holders, signer identities, signer independence, and treasury multisig address were not verifiable from the available sources. Timelock architecture exists, but the current delay is not verifiable as of October 13, 2026. Historical documentation described a 2-of-3 Safe behind a delayed multisig, initially 24 hours, with bypassable functions; that documentation is explicitly historical and cannot establish current settings.

Emergency/timelock bypass powers are present in the documented design. Legal/control entities: Terms identify Dolomite Ltd. (British Virgin Islands) as site/interface provider, Dolomite Foundation (Cayman Islands), and Leavitt Innovations, Inc.

(Delaware) as originating developer; governance documentation additionally references a Marshall Islands DAO entity and Leavitt Innovations as operator. Directors and registration numbers were not verifiable as of October 13, 2026. Terms use British Virgin Islands law and Road Town arbitration.

Voting concentration/top holders via Dune: Not verifiable as of October 13, 2026. On-chain verification was unavailable; no holder percentages or cross-chain exposure should be inferred. The same applies to current treasury signers and multisig threshold.

As of October 13, 2026: Entity and jurisdiction. Dolomite’s disclosed structure is multi-entity: Dolomite DAO LLC (Marshall Islands legal entity), Dolomite Ltd. (British Virgin Islands; interface provider and token issuer), Dolomite Foundation (Cayman Islands; steward), and Leavitt Innovations, Inc. (Delaware, U.S.; originating developer/contractor).

The Terms expressly group these entities for disclaimers, waivers, releases, and indemnities. Terms/restrictions. The Terms describe the protocol as decentralized and non-custodial: the interface claims no possession, custody, or control of user funds. Users must be legally capable, comply with applicable law, not be sanctioned or located in embargoed/restricted territories, and not use VPNs or anonymization tools to bypass restrictions. U.S. persons accessing “Non-OTC Contracts” must qualify as Eligible Contract Participants.

The interface also disclaims that loans are provided or endorsed by Dolomite entities. KYC/AML and sanctions. No public protocol-wide KYC/AML program, customer-identification policy, or Travel Rule implementation was verified. Access controls are primarily representations, geographic restrictions, sanctions screening obligations, and front-end controls; this does not establish regulatory compliance. The Terms prohibit sanctioned persons, while OFAC states that sanctions can also apply through ownership rules even without a listed entity. Classification. Dolomite characterizes its system as a non-custodial money market/DEX supporting lending, borrowing, margin trading, and other financial instruments.

Its Minerals terms state that Minerals are not securities or financial instruments; this is a contractual marketing position, not an independent regulator or court determination. DOLO’s legal classification is Not verifiable as of October 13, 2026. Warnings/enforcement/litigation. No regulator enforcement action, sanctions designation, or court case specifically against Dolomite, Dolomite DAO LLC, Dolomite Ltd., Dolomite Foundation, or Leavitt Innovations was identified in the reviewed public sources. This is not a clearance opinion.

Data-protection policy, GDPR/UK-GDPR role allocation, retention periods, and cross-border transfer safeguards were Not verifiable as of October 13, 2026. Risk view. The formal structure creates identifiable counterparties despite decentralized branding. The principal legal risk is cross-border regulatory exposure from U.S.-linked development/interface entities, derivatives or margin functionality, token issuance, and the gap between contractual restrictions and permissionless smart-contract access.

No exact GLEIF LEI record for 'Dolomite DAO LLC', 'Leavitt Innovations Inc'. OFAC SDN screening of 'Dolomite DAO LLC', 'Leavitt Innovations Inc': no match. SEC litigation and administrative release feeds: no mention.

Dolomite appears to rely on external stablecoins (for example USDC, USDT, DAI, and later USD1 on some markets) rather than issuing its own protocol stablecoin, so own_stablecoin is false. A depeg event is mentioned in the sources only for USDC in March 2023; the materials do not document any other stablecoin depeg tied to Dolomite, and no precise percent or count of Dolomite-specific depegs can be verified from the available sources, so those fields remain not verifiable as of 2026-10-13. The stablecoin used by the protocol was therefore at least one external stablecoin, but the exact stablecoin_ids cannot be confirmed from the gathered sources.

Dolomite’s principal risks are concentrated in oracle correctness, privileged administration, rapidly evolving modules/integrations, cross-chain dependencies, and liquidation/liquidity stress. Current chain-level exposure, TVL concentration, and administrator configuration are not verifiable as of October 13, 2026 because Dune/on-chain verification was unavailable. Contradiction: Dolomite documentation describes DOLO cross-chain operation using Chainlink CCIP, while its white paper references LayerZero; the effective production architecture is not verifiable as of October 13, 2026.

Dolomite’s top strengths are: (1) capital efficiency—it combines lending, borrowing, margin trading, and internal liquidity so one asset can serve multiple purposes at once; (2) modular architecture—a separate immutable core and mutable modules let the protocol add features without rewriting its safety-critical base; (3) collateral fidelity / preserved utility—users can keep assets productive while using them as collateral, rather than fully locking them away; (4) risk isolation—separate positions and isolated margin reduce contagion from one bad account to the rest of the system; and (5) multi-chain reach and extensibility—Dolomite is deployed across Arbitrum and Berachain, and its design supports broader integration and growth.