Felix CDP

Direct answer: Felix CDP is the CDP stablecoin component of the Felix Protocol, a Liquity V2‑style CDP and lending suite built natively on Hyperliquid L1, minting the feUSD synthetic dollar against crypto collateral. --- ### Protocol identification

• Name: Felix Protocol – Felix CDP (CDP market / feUSD).
• Category: CDP / overcollateralized synthetic dollar / lending protocol.
• Website: usefelix.xyz.
• Docs: Hosted on GitBook under usefelix.gitbook.io/docs.
• Chains: Hyperliquid L1 / Hyperliquid EVM only; DefiLlama lists chain as Hyperliquid L1.
• Native stablecoin: feUSD, an overcollateralized, USD‑pegged synthetic dollar minted via the Felix CDP market.
• Broader product suite: Felix comprises Felix CDP (feUSD) plus Felix Vanilla variable‑rate lending markets built on Morpho’s lending stack. Launch timing (high‑level):
• Felix is described as a leading native CDP on Hyperliquid with TVL >$150m by May 2025, implying launch in 2024–early 2025; exact genesis block date is Not verifiable as of 2026‑08‑28 from current open sources. ### Token & contracts (identification only)
• Stablecoin token: feUSD is the Felix CDP‑minted synthetic stablecoin.
• Protocol / CDP contracts:
• Felix’s public GitHub (felixprotocol/felix-contracts) contains the CDP smart contracts (“Felix CDP is a collateralized debt platform… Troves… feUSD”).
• Aggregators (DefiLlama “Felix CDP”, Hindenrank) reference Felix as a single CDP protocol on Hyperliquid L1, but do not surface canonical contract addresses directly in the snippets. Because on‑chain tools are unavailable in this run, main contract addresses and explorer verification status are: Not verifiable as of 2026‑08‑28. Any contract mapping from GitHub/docs would be an unverified marketing claim until checked on‑chain. ### Fork lineage & modifications
• Upstream: Multiple independent sources state Felix is a fork of Liquity V2.
• Key changes vs Liquity V2 (per open sources):
• Runs on Hyperliquid L1 rather than Ethereum mainnet.
• User‑chosen borrowing interest rate instead of purely algorithmic rates; liquidation priority depends on chosen rate.
• Expanded collateral set tailored to Hyperliquid: HYPE, UBTC/WBTC, kHYPE, wstHYPE, LSTs (stETH, rETH, etc.) depending on source/date.
• Integration with separate Vanilla money market built on Morpho for variable‑rate lending beyond feUSD. ### Audits & malicious‑fork history
• Public sources (Three Sigma case study, IQ.wiki, Hindenrank, DefiLlama) discuss architecture and risks but do not provide a direct, verifiable link to a formal smart‑contract audit report for Felix CDP or its Liquity‑based modifications in the retrieved snippets. Not verifiable as of 2026‑08‑28.
• No specific reports of malicious modifications in Felix itself or in directly related Liquity V2 forks are visible in the retrieved material. This absence is not proof of safety; it is simply Not verifiable as of 2026‑08‑28.

Felix CDP appears to be a real, functioning product portal rather than a pure landing page: the docs reference a live borrow flow on usefelix.xyz, including connect-wallet, deposit collateral, mint feUSD, and borrow actions, and Felix’s own materials point to a mainnet launch and a live mainnet link. An open API is present: Felix publishes a Felix REST API with a base URL at api.felix.net/v1, and its docs also describe a Web API for integrating Felix data with third-party systems. What is not fully verifiable from the available web evidence is the current live state of deposits/withdrawals in the UI, whether every documented flow is operational end-to-end, and whether there are broken links, fake metrics, or template-site signs; these require live UI inspection and/or on-chain verification, which is not available here. Not verifiable as of 2026-08-28. ## Maturity read

• Product maturity looks above landing-page stage because there are public docs, developer/API docs, and references to an active borrow app.
• The presence of APIs and developer documentation suggests some operational depth, not just marketing copy.
• UX quality, live deposit/withdrawal reliability, and any hidden template artifacts remain unverified as of 2026-08-28.

Focused code audit of Felix’s price-feed module for the Hyperliquid L1 Felix CDP system. Scope covered 198 non-comment, non-blank Solidity lines in src/PriceFeeds/, specifically KHYPEPriceFeed.sol and WSTHYPEPriceFeed.sol. The audit examined HYPE/LST/USD price-composition logic for canonical and market paths.

Fix status: both medium findings were addressed in commit #398ca73. The remediation explicitly composed HYPE/USDC × USDC/USD before combining with LST ratios, removing the hidden peg assumption, and right-sized deviation thresholds to the oracle’s real behavior. The source material indicates the issues were fixed in code, but the deployed-code coverage relative to current on-chain bytecode is not verifiable from the available web sources; per the bytecode-match note, this must be treated as not verifiable as of 2026-08-28.

Findings: 2 medium-severity issues and 1 informational issue were described in the available report material. Medium findings were: (1) inaccurate HYPE/USD due to an implicit USDC-peg dependency, and (2) incorrect deviation thresholds in KHYPE and WSTHYPE feeds. Informational finding: market price risk associated with kHYPE due to withdrawal restrictions. The source material does not indicate any critical or high findings for this review.

Felix CDP appears to have an active bug bounty program on Immunefi. The program is listed as live since 02 October 2025 and was announced by Immunefi on 3 October 2025 as a $100,000 bounty focused on Felix Protocol’s borrowing and lending products. Its stated scope includes the feUSD CDP Market, Stability Pool, Trove Manager, Price Feed, and Collateral Registry. The reward structure shown by Immunefi is Critical: $20,000–$100,000, High: $4,000–$10,000, Medium: up to $2,000, and Low: up to $1,000. The Immunefi program page also states that rewards for critical smart contract bugs are 10% of the funds directly affected, capped at $100,000, and that PoC is required plus KYC is required for payout processing. I could not verify any public disclosure of results such as total paid out, number of valid reports, or confirmed vulnerabilities fixed from the provided sources. Not verifiable as of 2026-08-28.

Felix CDP’s main dependency set appears to be internal to Hyperliquid rather than a large web of external DeFi protocols. Public coverage indicates the CDP accepts HYPE and UBTC as collateral, and also supports kHYPE and wstHYPE, which introduces dependencies on liquid-staking and wrapped-asset infrastructure, plus bridge/custody assumptions for bridged BTC exposure via UBTC. The clearest external risk is oracle dependence. Third-party analysis of Felix’s lending/CDP design says Felix uses RedStone as the primary pricing solution for crypto assets, and an audit report flagged price-feed issues for kHYPE and wstHYPE plus an implicit USDC/USD peg dependency in the HYPE/USD feed path. That creates manipulation and depeg risk if the oracle input, the USDC leg, or the liquid-staking asset pricing becomes unstable. Stablecoin exposure is material. feUSD is an overcollateralized synthetic stablecoin, so its value depends on the quality and liquidity of the underlying collateral and on liquidation mechanics, not on an off-chain reserve custodian. For stablecoin failure scenarios, the relevant risk is a collateral depeg or oracle failure rather than an issuer insolvency event. For LST/restaking exposure, publicly described collateral includes kHYPE and wstHYPE; audit findings note withdrawal restrictions and market-price risk for kHYPE, which can impair liquidation reliability during stress. That means a staking-layer or redemption bottleneck could transmit directly into CDP undercollateralization. I did not find verifiable evidence of meaningful RWA issuer/SPV, CEX market-maker, or custodian dependence for Felix CDP itself. The protocol materials and third-party coverage reviewed here do not establish such exposures clearly enough to verify them as of 2026-08-28. Failure / depeg / insolvency scenarios: oracle malfunction, USDC depeg in the HYPE/USD feed path, liquidation shortfalls for kHYPE/wstHYPE, or a hard depeg of the underlying collateral can all cause feUSD to break peg and leave the system reliant on stability-pool/liquidation capital to absorb losses.

Felix CDP is organized as a non-custodial protocol: users keep control of their own wallets, private keys, and recovery phrases, and the platform operator does not provide custodial services or manage user assets. The protocol’s collateral is held and managed by smart contracts on the blockchain rather than by Felix itself, and the docs explicitly say the platform does not store or manage user tokens.

No confirmed security incident, exploit, or user-loss event was found in the supplied sources for Felix CDP since launch. The only concrete third-party security material is a July 23–25, 2025 Three Sigma audit that reported two medium-severity pricing issues and one informational risk, all of which were remediated in commit #398ca73; this is a code-quality finding, not an incident. The protocol was also later reported to be shutting down its DEX/related markets in June 2026, but that source does not describe a hack or loss event.

Felix CDP’s key management is not described in the available sources in the sense of cryptographic key custody, multisig ownership, signer roles, or admin key controls. Based on the retrieved materials, the only clearly documented control/operations mechanism is protocol-side liquidation and redemption infrastructure, including an in-house liquidator bot, a partnership with Manhattan Research, and community-run bots to decentralize liquidation coverage. The protocol documentation also emphasizes that feUSD has no centralized custodian and relies on liquidators and redeemers for solvency, but that is not the same as key management and does not disclose how contract-admin or upgrade keys are organized.

Felix CDP is a lending/CDP product on Hyperliquid L1, which is a custom chain with non‑EVM architecture. As a result, standard Etherscan-style contract verification, proxy patterns and on-chain role inspection are Not verifiable as of 2026-08-28. ### 1. Contract addresses & verification

• Public sources reference Felix CDP as part of the Felix Protocol integrated natively in Hyperliquid’s perp/L1 environment, rather than as standalone verified EVM contracts.
• No canonical list of Felix CDP contract addresses with explorer-style verification was identified.
• Therefore, precise address set, source-code verification status, and ABI-level function lists are Not verifiable as of 2026-08-28. ### 2. Upgradeability / proxy / admin architecture Because Hyperliquid L1 is non-EVM, usual proxy/admin patterns (TransparentProxy, UUPS, ProxyAdmin, Timelock) do not directly apply, and there is no equivalent Dune/Etherscan decoded event trail.
• Proxy architecture, upgrade rights, and admin keys for Felix CDP are Not verifiable as of 2026-08-28. ### 3. Roles & privileged functions
• No independent technical documentation describing Felix CDP’s owner/admin/emergency roles, pause controls, oracle configuration rights, or fee/parameter setters was found outside protocol-controlled channels.
• Any role descriptions from Felix’s own docs would be unverified marketing claims under the given methodology. ### 4. Timelocks and governance
• There is no evidence from independent sources of an on-chain timelock protecting upgrades or parameter changes for Felix CDP on Hyperliquid.
• Governance model (multisig vs. DAO vs. company control) for CDP contracts specifically is Not verifiable as of 2026-08-28. ### 5. User exit conditions & worst-case key compromise Given the absence of verifiable low-level contract data:
• Whether users can always exit (repay and withdraw collateral) purely via immutable logic without admin cooperation is Not verifiable as of 2026-08-28.
• If Felix CDP is controlled by an upgradable program or privileged admin, worst case for key compromise would typically include:
• Forced parameter changes (collateral factors, interest, fees).
• Arbitrary liquidation or seizure of collateral balances.
• Freezing deposits/withdrawals via a global pause.
• Oracle manipulation leading to mass liquidations. These are *scenario-based inferences* drawn from common CDP designs on admin-controlled systems, not Felix-specific on-chain evidence. ### 6. Architecture map (high level) Based on general CDP patterns (not on-chain verified for Felix):
• Users: deposit collateral, mint stablecoins/borrow.
• Core CDP engine: tracks positions, collateralization, liquidations.
• Risk/oracle module: feeds prices and risk parameters.
• Admin/governance: can adjust parameters, potentially upgrade logic or pause. For Felix CDP on Hyperliquid, the concrete wiring of these components is Not verifiable as of 2026-08-28 under the available tooling and sources.

Based on currently accessible data, Felix CDP’s founders and core team are not publicly identified by name. All major ecosystem/analytics write‑ups describe the protocol’s design and performance but do not attribute it to specific individuals or a legal entity. As of 2026‑08‑28, detailed founder information is Not verifiable as of 2026‑08‑28. Public vs. anonymous; prior track record

• Ecosystem profiles (Hyperliquid guides, IQ.wiki, Hypurrcollective) describe Felix as a Liquity V2 fork and modular leverage suite on Hyperliquid L1, but none link to named founders, previous projects, or biographies.
• No credible sources surfaced tying Felix to a known DeFi team with disclosed prior hacks or major failures. Not verifiable as of 2026‑08‑28 whether the team is fully anonymous or pseudonymous, but the absence of names in independent coverage points toward *non‑disclosed* identity. Credibility signals
• Felix is consistently covered as a top native CDP/lending protocol on Hyperliquid, with substantial usage:
• Defiant notes ~$100m outstanding loans and $265m TVL, calling Felix the second‑largest protocol on Hyperliquid’s EVM.
• DefiLlama tracks Felix CDP TVL on Hyperliquid L1.
• Delphi Digital and Hypurrcollective research pieces treat Felix as a core Hyperliquid primitive, analyzing mechanics and risk, which indicates some institutional attention.
• Hyperion DeFi announced a formal partnership supplying 500,000 HYPE under a structured agreement, suggesting some level of due diligence by another institutional player. Real‑world footprint (office, jurisdiction, business status)
• None of the independent sources (media, analytics, ecosystem research) reference a registered company, physical office location, or jurisdiction.
• No evidence of onshore regulatory licensing or explicit offshore incorporation is reported. Not verifiable as of 2026‑08‑28. Reality check: real business vs. web‑front
• Activity metrics (TVL, loans, ecosystem integrations, research coverage) show Felix functions as an actively used DeFi financial primitive, not just a static website.
• However, the lack of disclosed founders, legal entity, and jurisdiction is a material governance/operational risk for institutional allocators, especially compared to fully doxxed, regulated borrowers. For institutional risk frameworks, Felix CDP should be classified as a high‑usage but team‑opaque protocol: strong on‑chain traction and ecosystem integration, weak verifiable information on founders, corporate structure, and real‑world presence as of 2026‑08‑28.

Felix CDP currently has a moderately positive but not blue‑chip reputation: active usage and multiple audits, with recognized risks around its novel design and Hyperliquid-specific exposures. No credible reports of fraud, rug pull, insolvency, or regulatory action were found as of 2026-08-28. Protocol & product reputation

• Felix is a Liquity V2 fork on Hyperliquid L1, providing an overcollateralized stablecoin feUSD via CDPs plus “Vanilla” lending markets.
• It is described by third parties as a leading native CDP protocol on Hyperliquid with substantial TVL and loan volume (e.g., >$100m outstanding loans, ~17% of Hyperliquid DeFi liquidity at one point). This indicates strong market adoption.
• Risk-ranking site Hindenrank rates Felix CDP Risk Grade B‑ / Value Grade D, explicitly calling out it as a Liquity V2 fork on Hyperliquid with user‑chosen interest rates. This reflects medium risk, limited value accrual from an institutional lens, not a “high-risk / likely to fail” label. Audits and security track record
• Felix’s CDP is built on Liquity V2, which has undergone multiple audits (Dedaub, ChainSecurity, Recon, Coinspect, Three Sigma) and formal verification by Certora; Felix-specific modifications are reported as being actively audited by Dedaub and Coinspect.
• A dedicated security review of the Felix Lending Protocol by Three Sigma (mid‑2025) found two medium‑severity issues in price-feed contracts and one informational issue around kHYPE market risk, with remediation recommendations. The presence of independent audits is a positive for reputation, though this does not guarantee safety. Team, investors, and transparency
• Public materials focus on the protocol’s architecture and usage rather than named founders or investors; founder identities and cap table are Not verifiable as of 2026-08-28 given current data.
• The GitHub repository for Felix contracts shows ongoing development activity, consistent with an actively maintained protocol. Sentiment, criticisms, and unresolved concerns
• Research outlets highlight Felix as an important Hyperliquid-native CDP, but risk-focused coverage flags design risks: user-chosen interest rates, reliance on Hyperliquid infrastructure and price feeds, and exposure to collateral-specific liquidity and oracle deviations.
• No credible allegations of fraud, rug pull, sanctions, or formal regulatory enforcement against Felix or feUSD are visible in independent sources as of 2026-08-28. Given the absence of on-chain verification in this run, all quantitative risk and usage assessments remain “Not verifiable as of 2026-08-28” against raw chain data and should be treated as aggregator-based interpretations rather than ground truth.

Felix CDP on Hyperliquid L1 is a collateralized debt position (CDP) stablecoin system: users deposit approved crypto collateral, mint feUSD, and can deploy that feUSD across the Hyperliquid ecosystem; the protocol is described by independent sources as a Liquity V2-style design, with users choosing their own interest rate rather than taking an algorithmic pool rate. The main assets in are HYPE and other Hyperliquid-native or listed collateral variants referenced by sources, including UBTC, kHYPE, wstHYPE, and some LSTs; the main assets out are feUSD, which is designed to stay redeemable for $1 of collateral. The yield engine is mostly organic, not clearly subsidized: borrowers pay ongoing interest and up-front/redemption/liquidation-related fees, while stability-pool depositors earn borrower interest and liquidation gains; Felix also has separate “Vanilla” lending markets and other Felix products, but those are distinct from CDP mechanics. Economically, the CDP is directional leverage rather than market-neutral: minting feUSD against collateral and swapping it increases exposure to the collateral’s upside/downside, while the protocol itself uses overcollateralization and liquidations to manage solvency. Lock-up is soft rather than fixed-term: collateral is locked while debt remains open, and withdrawals should depend on maintaining sufficient collateralization and repaying debt, but exact withdrawal gates, minimums, and product-specific limits are Not verifiable as of 2026-08-28 from the gathered sources. Fee detail from independent analytics is that Felix CDP revenue is composed of interest, redemption fees, and liquidation profit; however, exact fee rates and protocol take are Not verifiable as of 2026-08-28. For TVL, DeFiLlama shows Felix CDP at about $201.84m in USD terms and about $237.56m when denominated in HYPE, but the gathered sources do not provide a Dune-backed on-chain reconciliation, so the on-chain vs. aggregator gap is Not verifiable as of 2026-08-28. APY history appears variable rather than fixed: independent writeups describe borrower-chosen rates and floating lender yields, implying volatility tied to utilization, collateral mix, and liquidation activity; a precise APY time series and sustainability analysis are Not verifiable as of 2026-08-28.

Not verifiable as of 2026-08-28. The available web results do not provide a reliable, on-chain-verifiable treasury/reserve breakdown for Felix CDP on Hyperliquid L1, and Dune/on-chain queries are unavailable in this run. The protocol’s own or partner materials mention scale—such as “tens of millions of feUSD debt” and “over $250 million in deposited collateral”—but those are not independently verifiable reserve or custody disclosures and should be treated as unverified marketing claims. DefiLlama only gives protocol TVL snapshots, which are not sufficient to determine reserve assets, treasury addresses, control, custody structure, or attestations. I also found conflicting TVL figures across sources and dates, underscoring that any treasury size estimate here would be unreliable without raw on-chain checks.

For Felix CDP on Hyperliquid L1, a Bitcoin drop below $10,000 would primarily translate into a severe collateral-liquidity shock: borrowers against BTC would be pushed toward liquidation, Stability Pool capacity would be tested, and feUSD could face secondary stress if liquidations and redemptions accelerate. The protocol’s documented risk model says borrowers face liquidation when collateral value falls below threshold, and Stability Pool depositors absorb liquidated debt in exchange for volatile collateral. What is not verifiable as of 2026-08-28 is the exact protocol-wide impact under this BTC scenario—such as the share of BTC-backed positions that would liquidate, remaining overcollateralization after liquidations, or whether feUSD would stay near peg—because no on-chain data or live risk dashboard was provided here. The key failure modes to watch are:

• Liquidation cascade if BTC-backed CDPs cross liquidation thresholds simultaneously.
• Stability Pool depletion if liquidatable debt exceeds absorber capacity, forcing more disorderly unwinds.
• feUSD peg pressure if market participants expect broader liquidations and seek to redeem or sell feUSD. Felix’s own documentation indicates feUSD redemptions and liquidations are designed to support the peg, but that mechanism can concentrate stress on the lowest-collateralized positions during sharp collateral declines. So, the stress answer is: Bitcoin below $10,000 would likely be a worst-case collateral stress event for Felix CDP, with liquidation and peg-risk transmission channels, but the magnitude of protocol damage is Not verifiable as of 2026-08-28.

Felix CDP is designed as an over-collateralized borrowing system, so a 20% depeg in the largest collateral would reduce the dollar value of that collateral and directly weaken borrower collateralization ratios; if positions fall below their liquidation threshold, they can be liquidated and their collateral seized to repay debt plus a bonus. Felix’s documentation states that liquidation is triggered when collateral value falls below the asset-specific threshold, and that collateral value declines are a valid liquidation path. The specific systemwide loss from a 20% depeg is not verifiable as of 2026-08-28 because the on-chain collateral mix, largest-collateral share, and live debt distribution cannot be confirmed here without Dune/on-chain checks. The only collateral composition visible in the provided results is that Felix CDP accepts HYPE, UBTC, kHYPE, and wstHYPE, but those sources do not provide a current TVL/exposure split for Hyperliquid L1.

For Felix CDP on Hyperliquid L1, a top-counterparty insolvency stress path is not verifiable as of 2026-08-28 from the available non-onchain sources, so the exact contract-level loss waterfall, insurer/backstop, and compensation mechanics cannot be confirmed here. The most defensible externally sourced failure mode is a liquidation/reflexivity spiral: if the stressed counterparty is also the dominant collateral source or borrower cluster, forced liquidations can deepen price declines and propagate losses to other borrowers and liquidators rather than to a separate guarantor. Expected loss path: the insolvent counterparty would first fail to meet CDP obligations, then its undercollateralized positions would be liquidated; any shortfall would be realized against protocol reserves only if Felix explicitly maintains such reserves, otherwise the loss remains with the affected users/positions. I cannot verify a dedicated insurance fund, mutualized backstop, or lender-of-last-resort mechanism from the provided sources, so who absorbs it is not verifiable as of 2026-08-28. Compensation: absent a documented protection layer, compensation would typically be limited to whatever collateral is recovered through liquidation and any protocol-defined surplus distribution; any claim of full make-whole coverage is unverified marketing claim unless independently documented. The available source does not establish a compensator, so the safest statement is that recovery, if any, flows from liquidation proceeds back through the protocol’s accounting rules, while deficits are borne by the insolvent account’s creditors/users. Impact path through smart contracts: the likely sequence is oracle price update → collateral ratio breach → liquidation trigger → collateral seizure/auction/forced sale → repayment to debt ledger → any residual bad debt or shortfall remains unmatched. Because contract code and deployed addresses were not independently verified in the provided results, the exact module names and call sequence are Not verifiable as of 2026-08-28.

For the “committed fraud by the DAO or owners” stress case, I found no verifiable evidence that Felix CDP has a DAO, and the available terms state that market creation and parameter adjustments are controlled by the Platform Operator, with no DAO governance in place. That means the relevant fraud scenario is operator/owner misconduct, not DAO fraud. On the evidence available, Not verifiable as of 2026-08-28 whether Felix’s owners/operator have committed fraud. The strongest public sources instead describe a non-custodial product with user-controlled positions, plus audited/oracle-related design issues and concentration risks, but not allegations or findings of fraud. For a stress analysis, the main impact of a proven owner/operator fraud event would be governance/control abuse: malicious parameter changes, selective market modifications, misleading disclosures, or compromised administrative control. Because Felix’s market and parameter control are centralized in the Platform Operator, such an event could directly affect redemption logic, liquidation settings, and user trust. However, the specific harm magnitude is Not verifiable as of 2026-08-28 from the available sources. Separately, independent reviews highlight technical and economic stress risks that are not fraud: oracle deviation-threshold issues, an implicit USDC peg assumption in price paths, and kHYPE withdrawal/market-price risk during stress. These are relevant because a fraud scenario could exploit or worsen those existing weaknesses, but the sources do not show that fraud has occurred.

Felix CDP’s primary yield source is not verifiable as negative over the last 30 days from the provided results. The available sources describe Felix as an over-collateralized lending/CDP system with Stability Pool mechanics and liquidation/redemption risks, but they do not provide a 30-day yield series or a confirmed negative yield calculation for the primary yield source. What *is* supported is that Felix’s yield/risk profile is stress-sensitive: borrowers face liquidation and redemption risk, and Stability Pool depositors can receive volatile collateral when liquidations occur. Independent analysis also highlights a HYPE-collateral reflexivity spiral under sharp price declines, which could worsen realized outcomes for some participants during stress. The most relevant verification gap is the yield metric itself: there is no on-chain or independently reported 30-day primary-yield figure in the supplied material, so the requested negative-yield stress condition is Not verifiable as of 2026-08-28.

Felix CDP’s legal posture is only partly verifiable from public materials. The available terms identify a “Platform Operator” but do not clearly disclose a governing corporate entity or a specific jurisdiction for the CDP product itself, so the entity/jurisdiction remains Not verifiable as of 2026-08-28. The terms do state that the CDP is a fork of Liquity licensed to the operator, and that HyperLiquid is merely the third-party blockchain infrastructure used for perpetual markets, not the legal operator of the CDP product. On restrictions, Felix’s terms define “Restricted Users” to include persons or entities in restricted jurisdictions, sanctioned/high-risk screened wallets, and anyone acting for their benefit. The same terms place AML/CTF and sanctions compliance responsibility on users, and say failure may expose users to legal or regulatory consequences. Felix’s privacy policy also states it is non-custodial execution infrastructure and references GDPR/CCPA-style rights, but that is a privacy-policy statement rather than a full legal structure disclosure. No independent evidence found in the retrieved sources shows a court case, enforcement action, or sanctions designation against Felix CDP itself; those items are Not verifiable as of 2026-08-28. The strongest legal-risk signal is therefore the gap between the protocol’s operational design and its legal disclosure: the protocol appears to rely on user-side compliance while offering limited public clarity on the operating entity, venue, and regulatory perimeter. Sources: Felix Terms & Conditions; Felix Privacy Policy.

Yes — the stablecoin used by Felix CDP is feUSD, and the web results indicate it has depegged before. One source reports an all-time low of $0.7059, which implies a maximum downside depeg of about 29.4% below $1.00. However, the results do not provide a complete time series or an exact count of distinct depeg events, so the number of times and the date of the last depeg are Not verifiable as of 2026-08-28 from the available sources. The strongest supported statement is that feUSD has at least one documented depeg event, but the full frequency and recency cannot be confirmed here.

The top 5 risks for Felix CDP on Hyperliquid L1 are: 1) Collateral concentration / reflexivity: Felix relies heavily on HYPE and other Hyperliquid-native assets, so a sharp HYPE drawdown can trigger liquidations that further depress collateral prices in a negative spiral. 2) Oracle and pricing risk: its CDP system depends on correct collateral valuation, and the Three Sigma review flagged issues such as implicit USDC/USD dependence and multi-hop price-feed/decimal-scaling errors that can misprice collateral. 3) Peg and redemption risk: if feUSD trades below $1, redemption and liquidation mechanics are stressed, which can increase forced closures and destabilize the peg. 4) Chain / infrastructure dependence: Felix is built natively on Hyperliquid, so L1 outages, validator/sequencer issues, or broader ecosystem stress directly propagate to the protocol. 5) Rapid-growth / parameter-calibration risk: Felix’s fast deposit growth means core risk parameters may not yet be fully battle-tested across a full market cycle, increasing the chance of under-calibrated liquidation, rate, or redemption behavior.

Felix CDP’s top strengths are: 1) a proven design: it is a Liquity V2-style / Liquity V2 fork, which supports confidence in the core CDP mechanics; 2) native stablecoin issuance: users can mint the Hyperliquid-native stablecoin feUSD against approved collateral; 3) flexible borrowing terms: borrowers can set their own interest rates, giving them more control over financing costs; 4) high leverage potential: borrowing is constrained mainly by collateral mint caps rather than lender-side liquidity, which can support larger positions; and 5) ecosystem fit and composability: Felix is built on Hyperliquid L1, where its CDP and lending products are positioned as part of a broader on-chain trading and liquidity stack.