stHYPE

Tier 2

Security Analysis Report

TVL

$183.23M

Sections

7

report sections

Citations

341

evidence references

Version

v1

10/29/2026

5.3

Moderate Risk

Safety Score · out of 10

📋

Executive Summary

5.3/10

Executive Summary: Sthype

stHYPE is a liquid staking token for HYPE on Hyperliquid, with about USD 183M in TVL (money deposited), all on Hyperliquid L1. Its staking contracts have been reviewed by several audit firms, and no hack is recorded. The main concerns are team-held control, slow or discounted exits under stress, and no guaranteed loss cover. Most control and exit facts come from the protocol's own documentation (source reliability: unverified publisher), and no stHYPE contract on HyperEVM was read on-chain, so confidence is moderate to low.

Safety Rating

Overall Score: 5.3/10 Dimension scores: Code security 6.0 · Economic security 5.5 · Team transparency 3.5 · Governance 4.7 · Track record 6.1

Scores by dimension, points out of 10 (points)
Code security
6.0 points
Economic security
5.5 points
Team transparency
3.5 points
Governance
4.7 points
Track record
6.1 points

The rating reflects strong audit activity on the staking contracts, held back by the fact that the audited code is in a private repository and cannot be matched to the deployed contracts. It also reflects an anonymous team and a multisig that cannot be checked on-chain.

Scores by Dimension

DimensionKey Finding
Code SecurityFour firms reviewed the staking contracts, but the audited repository is private, so deployed code was not compared with it (a measurement gap, not a finding that the code differs)
Economic SecurityExits can be slow, no guaranteed insolvency backstop is documented, and holder concentration and DEX depth look weak on unverified, partly dated data
Team TransparencyNo official team page names an accountable person; named founders and the Panama operating company rest on unverified publications
GovernanceA multisig, not token voting, controls the protocol; the docs describe a 48-hour upgrade delay, but the official registry does not confirm the timelock ownership for the core proxies, and no delay is documented for role or parameter changes
Track RecordNo incident is recorded, but evidence of behaviour under stress is weak and the operating history is short

What This Protocol Does

stHYPE lets users deposit HYPE and receive stHYPE, a token that represents their stake and keeps earning staking rewards. Per the Valantis documentation, the protocol stakes the deposited HYPE on HyperCore (Hyperliquid's native layer) and also offers a trading-fee-discount product (Valantis Prime) and a restaking pool. These mechanics are Valantis's own descriptions from a documentation site whose publisher is not verified (source reliability: unverified publisher), and they are not independently confirmed. DeFiLlama records about USD 183.2M TVL (2026-10-29) for stHYPE on Hyperliquid L1 only. The STEX AMM and the wider Valantis AMM stack are separate Valantis products and are context only.

Key Strengths

  • Audit depth on the staking contracts (stHYPE): seven reports from four firms over 20 months, with remediation reviews, and all critical and high findings fixed per the audits. Several of these audits sit on hosts not verified as auditor-controlled, which lowers confidence in provenance but not coverage credit. Audits of the older Valantis AMM products add nothing here.
  • No leverage or rehypothecation in the core token (stHYPE): the docs say stake accounts can only stake to HyperCore. The docs (source reliability: unverified publisher) and a third-party review are the only sources, so this is unconfirmed by an independent source.
  • Reduced admin power (stHYPE, wstHYPE): the docs say the admin's ability to block individual addresses was removed on 2026-04-10, and the docs describe a 48-hour upgrade timelock. Neither is confirmed on-chain, and the registry does not confirm the timelock for the core proxies.
  • Rebase safety bounds (stHYPE): the code Pashov reviewed in October 2025 reverts the rebase on an excessive rate change or supply drop. Whether the deployed code has the same bounds is not established, because deployed code was not compared with the audited versions.
  • Clean incident record: the DefiLlama hacks snapshot holds no records for the protocol or its family. An empty list does not prove none occurred.

Key Risks

  • 1. Exit stress, delayed or discounted exits (stHYPE and wstHYPE holders). Magnitude: the delay applies to all USD 183M; likelihood: medium for a visible discount or queue delay, low for a prolonged freeze. If the instant-redemption queue empties, later exits depend on slower accounts. Unverified explorer data shows the top ten stHYPE holders owning about 90.7%, with unlabelled holders that may be contracts. Valantis Prime traders face a separate limit: per the docs (source reliability: unverified publisher), if the protocol ends a Prime account, balances stay inaccessible for at least seven days, with no guaranteed maximum duration. A Yearn risk report from May 2026 (source reliability: unverified publisher) puts DEX liquidity at about USD 6.95M; this is an unconfirmed figure.
  • 2. Admin multisig compromise or misuse (stHYPE, wstHYPE, OverseerV1, restaking escrow). Magnitude: up to the full USD 183M in the worst case; likelihood: low. Four of six unnamed signers can change roles and parameters, and upgrade after the documented 48-hour delay (not confirmed for the core proxies). Holders have no veto, and a full exit can take longer than the delay. The threshold, signers and timelock are documentation statements and are not confirmed on-chain for HyperEVM.
  • 3. Slashing, deployer exposure and socialised losses (stHYPE, wstHYPE, restaked positions). Magnitude: bounded by stake in the affected accounts, which is not measured; likelihood: low. For HIP-3 deployers (builder-deployed perpetual markets), penalties can reach the entire stake. Which HIP-3 or HIP-4 rules apply to the only active deployer account, the USDe quote-asset module, is not established, and base-staking slashing was not yet enabled at Pashov's October 2025 review. Slashed HYPE is burned, with no guaranteed backstop. A Staking Rewards review (source reliability: unverified publisher) reports the treasury at under 0.25% of TVL; this is unconfirmed.
  • Single-keeper pause on the restaking escrow (restakers in StakedHypeEscrow). Magnitude: moderate, limited to restaked funds, whose size is not measured; likelihood: low. The docs (source reliability: unverified publisher) say one keeper holding MANAGER_ROLE can call setPaused, which blocks deposits, requests, cancellations and execution with no maximum duration. Restakers cannot exit while the pause lasts. This is an unconfirmed report, and the escrow audit does not mention the pause.
  • DeFi integration risks (wstHYPE borrowers, lenders and integrators outside this record): the docs (source reliability: unverified publisher) warn that the dual-token model (stHYPE and wstHYPE) introduces a significant vulnerability for certain DeFi integrations and say mitigation is simple. The reports do not establish the mechanism or the affected integrations, report no exploit, and do not link this warning to the liquidation scenario below. Separately, liquidation stress for wstHYPE collateral: a Yearn risk report from May 2026 (source reliability: unverified publisher) counts about USD 65M of wstHYPE used as collateral against about USD 690K of wstHYPE DEX liquidity. These are unconfirmed, five-month-old figures, and the lender data is not in the sources.

Critical Findings

No critical findings identified. One high-severity item is open, but it concerns the separate Valantis HOT AMM product, not stHYPE: the signer is a single point of failure, a High (estimated, an AI estimate; the audit states no rating), acknowledged by the team. Two high findings on the staking contracts (Guardian, November 2025 and January 2026) are fixed per the audits and were not re-checked.

  • Two medium findings on the stHYPE Overseer from the Guardian review of 2026-01-05 are acknowledged, not fixed. One says getTotalBalance leaves out a deprecated interim address, which can cause false slashing or block mint and redeem. The other, "Min Slash Threshold Skews Losses", means small slash losses fall on current holders. The audited repository code was not available, so these were not checked against code.

Recommendation

Assessment: stHYPE suits risk-tolerant HYPE holders who accept team-held control, slow exits under stress and no guaranteed loss cover in return for staking yield. This is a judgement based on the findings cited below, not a recommendation made by the worker reports. The audit record on the staking contracts is a real strength. The weak points are an anonymous team, a multisig that cannot be checked on-chain, and thin secondary liquidity. Assessment: users with low risk tolerance, or who may need to exit large amounts quickly, should wait until the multisig, timelock and roles are confirmed on-chain and the audited code is tied to the deployed contracts (the audited repository is private, so this is a measurement gap, not a finding that the code differs), because the cross-risk report lists these as missing facts. Looped wstHYPE positions in lending markets carry extra liquidation risk.

Report Sections

  • Protocol Overview
  • Code Security
  • Economic Security
  • Social & Team
  • Governance
  • Cross-Risk Analysis

44 citations · click markers to view evidence

🏗

Protocol Overview

Protocol Overview: Sthype

What It Does

stHYPE is a liquid staking protocol for HYPE, the token of the Hyperliquid network. You deposit HYPE and receive stHYPE, a token that represents your stake and keeps earning staking rewards. The protocol stakes the deposited HYPE on HyperCore (Hyperliquid's native layer) and also offers a trading-fee-discount product, Valantis Prime, and a restaking pool. The DeFiLlama record covers only stHYPE on Hyperliquid L1, with about $183.2M TVL in the latest snapshot (2026-10-29). Mechanics, controls and risk statements come from the Valantis documentation site, whose publisher is not verified (source reliability: unverified publisher). They are Valantis's own descriptions, not independently confirmed. The stHYPE contracts could not be read on-chain for this report, and their audited source code is not public.

  • Category: liquid staking (DeFiLlama category "Liquid Staking") .
  • Product set under this record: the stHYPE token and its OverseerV1 staking manager, the yield-bearing wstHYPE interface to the same share data (users hold both tokens on HyperEVM, with no wrapping required) , HyperCore stake accounts, the restaking escrow, and Valantis Prime accounts. The STEX AMM (a pool built for LSTs, which are liquid staking tokens) and the Valantis AMM stack are separate Valantis products. They are context here, not part of the stHYPE contract set.
  • Target users: HYPE holders who want staking yield while keeping their capital usable . Traders who want lower Hyperliquid trading fees use Prime. Restakers want extra yield from deployer support.
  • Why users stake: staking on Hyperliquid also lowers trading fees, lowers trading latency, and unlocks HIP-3 perps market deployments . stHYPE makes those uses liquid.

How It Works

stHYPE turns HYPE into a tradable receipt token while the real HYPE stays staked on HyperCore. The mechanics below are Valantis's own description from its documentation site (source reliability: unverified publisher). The protocol describes itself as "a single unified Staking Protocol that coordinates several Stake Accounts on HyperCore" .

User flow

  1. The user gets HYPE onto HyperEVM (chain ID 999). The user buys spot HYPE on Hyperliquid and transfers it to HyperEVM .
  2. The user connects a wallet at the Valantis app, enters an amount, optionally picks a validator through the community codes program, and clicks Stake . The deposit goes through the OverseerV1 proxy, which mints stHYPE from the stHYPE token contract .
  3. The Overseer's manager role moves deposits into staking modules (HyperCore stake accounts), which stake to HyperCore. New deposits first settle queued withdrawals before anything is staked, and the contracts enforce this . These accounts "can only stake to HyperCore. None of them lend or rehypothecate."
  4. Rewards reach holders through rebases. The REBASER_ROLE keeper calls rebase on the Overseer . Staking, unstaking and rebasing are handled by permissioned keeper functions . Holder balances then grow. wstHYPE instead shows rising value per token: "1 stHYPE = 1 HYPE" while "1 wstHYPE > 1 HYPE" . Both are views of the same share data.
  5. To exit, the user calls burn, burnAndRedeemIfPossible, or redeem, or swaps stHYPE on a secondary market . A normal native withdrawal targets about 7 days with no fee . An instant-redemption queue and an AMM swap give faster exits when liquidity exists .
  6. Optional: the user deposits into the restaking pool, StakedHypeEscrow (SHE). The deposit function transfers wstHYPE , and the position is read through the shared stHYPE share accounting . Exit requires a withdrawal request, a cooldown, then a claim that pays out wstHYPE .
stHYPE deposit, staking and exit flowOpen full view →
ACTORHYPE holderCONTRACTOverseerV1 (Overseer)TOKENstHYPE tokenCONTRACTHyperCore stakeaccountsROLERebaser keeperdeposits HYPE2mints stHYPE2stakes via modules3calls rebase4
  • (actor)
  • (contract)
  • (token)
  • (contract)
  • (role)
  • step 2flow
  • step 2flow
  • step 3flow
  • step 4control

Stake account types

  • Fast-withdrawal accounts (5): hold most assets. HyperCore allows five concurrent unstake requests per account, so five accounts give 25 concurrent unstakes .
  • Deployer accounts (1 active): support Hyperliquid deployers to earn above native staking yield, with slow withdrawals and timelocked movement of funds .
  • Prime stake accounts (more than 250): each trader gets an isolated account, with an equal amount of the protocol's native stake moved in . The protocol can end a Prime account through a notice-then-shutdown path. During shutdown the trader loses HyperCore access and the balances stay inaccessible for at least 7 days .

Key contracts (HyperEVM mainnet, source-provided)

ContractAddressRole
stHYPE (proxy)0xfFaa4a3D97fE9107Cef8a3F48c069F577Ff76cC1Rebasing receipt token
OverseerV1 (proxy)0xB96f07367e69e86d6e9C3F29215885104813eeAEMint, burn, queue and staking-module manager
wstHYPE (proxy)0x94e8396e0869c9F2200760aF0621aFd240E1CF38Yield-bearing form
StakedHypeEscrow0x0447d0339c82a8EFe01994b7D1981bf5c517da0eRestaking pool

Addresses come from the protocol docs and the restaking page . No on-chain match to source code was available for them.

Yield Sources

stHYPE yield comes mainly from Hyperliquid staking rewards, with optional extra yield from restaking. Restaking rewards can take several forms, so their sustainability varies by allocation.

  • Base yield: "stHYPE rewards primarily track staking-related yield from HyperCore staking activity."
  • Protocol fee: a standard 10% protocol fee is already included in the app's APY, and there is no deposit or withdrawal fee .
  • Extra yield (restaking): deployers pay for stake support, and restakers share a pool of rewards . The documented rewards table lists USDe Quote Asset at 10 HYPE per day (active), HIP-4 deployer at 40 HYPE per day (pending liquidity, not live), and a market-maker discount stream at an estimated 20–50 HYPE per day (coming soon) . The docs say these numbers can change.
  • Unallocated extra yield: at Valantis' discretion it goes to all holders through rebases or is kept by the protocol .
  • Prime fee: trades through the Prime UI pay Valantis 30% of the trading-fee savings from staking .
  • No fixed return: rebases and premium streams "Neither guarantees a fixed HYPE return."
  • Historical yield ranges: not available in the sources. Current APY is shown only in the app.

Assessment: base yield is staking-driven and does not rely on a stHYPE emission schedule that the sources describe. Deployer rewards "can be HYPE premiums, token incentives, points, or revenue share" . The mix per allocation was not checked. The listed restaking premiums are small and one stream is still not live, so most holders earn mainly base staking yield. The Valantis Points programme page exists in the sources but was not analysed as a yield source.

Dependencies & Integrations

stHYPE depends almost entirely on Hyperliquid itself.

  • Hyperliquid (HyperCore and HyperEVM): rewards, unbonding limits and slashing are set by Hyperliquid. "At a high level, rewards and base unbonding constraints are determined by Hyperliquid network mechanics." Stake accounts use CoreWriter, which are contracts that send actions to HyperCore.
  • Validators: the user can pick a validator through the community codes program. Slashing applies to validators that fail protocol-defined duties.
  • Deployers and quote assets (Ethena USDe): the active deployer account is the USDe quote-asset module. Ethena's HyENA HIP-3 allocation is deprecated and the proposed HIP-4 deployer is not live .
  • Oracles: the sources do not describe a price oracle for core stHYPE mint and redeem. Mint and redeem rely on the protocol's own share accounting and rebases.
  • Chain: DeFiLlama attributes all TVL to Hyperliquid L1 . Bridge dependency: users reach HyperEVM by transferring from HyperCore balances, and USDC deposits reach Hyperliquid through Arbitrum or other routes .
  • Secondary liquidity: the STEX AMM (separate Valantis product) has an stHYPE/WHYPE pool . The Valantis swap router is universal to Valantis pools on HyperEVM.

TVL & Adoption

DeFiLlama reports $183.2M TVL for stHYPE in a snapshot collected 2026-10-29 .

  • Chain distribution: 100% Hyperliquid L1 . The dependency data also lists only Hyperliquid L1.
  • Historical trend: no dated historical series is in the sources.
  • Adoption signal: the docs report more than 250 Prime stake accounts .
  • Hacks: the captured DeFiLlama hacks list has no records for this protocol.
  • Chains with no current TVL: none are listed.

Technical Architecture

According to the Valantis documentation (source reliability: unverified publisher), a multisig controls upgrades and operations, a 48-hour timelock delays proxy upgrades, and no holder veto exists. These statements are unconfirmed until the contracts are read on-chain.

Who controls what (documented, not confirmed on-chain)

The Valantis documentation (registry page dated 2026-10-16; source reliability: unverified publisher) describes these controls:

  • Admin multisig: DEFAULT_ADMIN_ROLE on stHYPE, Overseer and wstHYPE is held by 0x97dEe0eA4Ca10560f260A0f6f45bdC128A1D51f9, with a 4/6 threshold, updated 2026-04-07 . This role administers roles, staking-module and parameter controls, and pause controls .
  • Upgrades: proxy implementation upgrades are multisig-administered and "subject to a 48-hour on-chain timelock" . The timelock went live on 2026-04-10 for the proxy contracts and staking modules .
  • Keepers: staking, unstaking and rebasing run through permissioned keepers that cannot act outside contract bounds and never custody funds . The MANAGER_ROLE runs module operations, REBASER_ROLE calls rebase, and FEE_RECIPIENT_ROLE calls claimFee.
  • Admin transfer delay: for the core contracts the registry describes, the default-admin transfer delay is 0 in current deployments, separate from the 48-hour upgrade timelock . The restaking escrow has a separately documented two-day default-admin transfer delay, which does not delay its setters or operational-role changes .
  • Prime shutdown control: stHYPE can start a timed shutdown if commercial terms end . The shutdown controls change only after a time lock passes .
  • Removed power: on 2026-04-10 the admin's ability to block individual addresses from transferring stHYPE and wstHYPE was removed .
  • Voting: no on-chain veto exists. A Lido-style dual-governance model is a roadmap direction after Valantis Governance is adopted in 2026 .
  • Restaking escrow (SHE): a keeper holds MANAGER_ROLE, which can pause and rescue unrelated tokens. The admin multisig can set cooldown (1 second to 180 days) and reward duration (1 second to 7 days). Upgrades go through a timelock with a 48-hour minimum delay . The SHE pause has no maximum duration, and it blocks deposits, requests, cancellations and execution together .

On-chain check: the measured on-chain control data in the workspace covers the Valantis AMM contracts on Ethereum, Base and Arbitrum, not the stHYPE contracts on HyperEVM. The multisig signers, threshold, timelock and roles above are therefore the docs' statement, not confirmed on-chain. The docs date them 2026-04-07 to 2026-10-16.

Emergency mechanisms

  • Automatic checks: a pending slash blocks mint and makes redeem non-redeemable. Rebase reverts on an excessive APR or supply drop (AprTooHigh, SupplyDecreaseTooHigh) .
  • Operator pauses: off-chain alerts do not auto-submit pauses. Operators call pauseRedemption, pauseBurns or both, bounded to more than 0 and at most 14 days . The Overseer pause duration is set to 7 days at V3 initialization .
  • Parameters: governance can change the APR threshold (set to 500 bps in the changelog), pause duration, cooldowns and staking-module allocation .
  • No backstop: "There is currently no guaranteed insolvency backstop for user losses."

Upgradability

Core contracts sit behind proxies (the docs label the addresses "Proxy"), with ProxyAdmin ownership under a timelock the multisig runs as proposer, canceller and executor .

Who can change stHYPE contracts, per the docsOpen full view →
GOVERNANCEAdmin multisig 4/6CONTRACTUpgrade timelock 48hCONTRACTOverseerV1 proxyROLERebaser keeperCONTRACTStakedHypeEscrowproposes upgradesupgrades after 48hgrants rolescalls rebaseupgrades after 48h
  • (governance)
  • (contract)
  • (contract)
  • (role)
  • (contract)
  • control
  • control
  • control
  • control
  • control

Key Risks (Protocol-Level)

User funds sit as HYPE staked in HyperCore stake accounts controlled by protocol contracts. The main risks are slashing, withdrawal delays under stress, admin control and unreviewable code.

  • Slashing and deployer exposure (impact: high; likelihood not estimated): validators can be slashed, and deployer and quote-asset accounts add their own slashing surfaces . Deployer recall periods range from 30 to 183 days . Likelihood is not estimated because the sources give no slashing history and no compliance record for the active USDe deployer account. Historical validator slashing events and that account's market-compliance history would decide it. The docs state there is no guaranteed insolvency backstop for user losses .
  • Withdrawal liquidity: prolonged withdrawals remain possible if requests exceed what standard accounts and new deposits absorb . Secondary-market stHYPE can trade below HYPE in that case.
  • Admin and upgrade control (medium): per the Valantis documentation, a 4/6 multisig can upgrade proxies after 48 hours, and no holder veto exists. Neither is confirmed on-chain. Funds are not held by the keepers, but the admin multisig can change implementations.
  • Prime shutdown (traders): a Prime account can become inaccessible for at least 7 days during shutdown, with no guaranteed maximum duration .
  • Restaking exposure: SHE withdrawals carry a cooldown, documented as 14 days when checked on 18 October 2026 . The admin can change it for future requests, and the current setting was not independently checked. Reserved shares keep slashing exposure meanwhile. SHE can be paused with no maximum duration .
  • Smart contract risk: the docs state that smart contracts can still contain vulnerabilities despite audits and testing . There is no active public bug bounty .
  • Complexity: the system has five fast-withdrawal accounts, a deployer module, more than 250 Prime accounts, two token views (stHYPE and wstHYPE), a restaking escrow, and an off-chain keeper layer. The Valantis documentation (source reliability: unverified publisher) calls the stHYPE and wstHYPE interfaces a novel token mechanism that should be reviewed carefully . It also warns that the dual-token model introduces a significant vulnerability for certain DeFi integrations and says mitigation is simple . The docs add that a transfer, mint or burn of stHYPE performs the same action for wstHYPE .
  • Dependence on Hyperliquid: a change in Hyperliquid staking, slashing or CoreWriter behaviour would affect all stHYPE accounts. The docs state that "Hyperliquid and stHYPE are evolving. Features and use cases may change as the network develops."

Audit coverage by product

Audits in this workspace for stHYPE contracts, by auditor and date:

  • Pashov Audit Group, Oct 13–16, 2025: Overseer, HyperCoreUtils, HyperCoreStakingModule .
  • Pashov Audit Group, Nov 10–11, 2025: StakingModuleExternalManagement.
  • GuardianAudits, Nov 2025: StakingModuleExternalManagement. GuardianAudits, Jan 2026: Overseer and its V3 upgrade test.
  • Reviews dated Feb 2025, Feb 2026, Apr 2026 and Oct 2026: the Apr 2026 review lists StakeAccountManager and StakingModuleExternalAccount. The Oct 2026 report is on the restaking escrow. The Feb 2025 and Feb 2026 scopes could not be determined from the available material.
  • The docs list Three Sigma (Feb 2025), Obsidian (Apr 2026) and others on the stHYPE audit page . Several copies are hosted on a GitHub path whose publisher is not verified as the auditor.

Audits of Valantis Core, HOT, the swap router and STEX (Hexens, Statemind, Sherlock and others) cover the AMM stack, not the stHYPE token, Overseer or Prime. They are not review depth for stHYPE. The audited stHYPE source repository is private, so the audited code could not be compared with the deployed contracts. Treat that comparison as not performed.

Data Completeness

  • Available: a DeFiLlama TVL snapshot (2026-10-29), extensive protocol docs (mechanics, withdrawals, restaking, Prime, governance registry dated 2026-10-16, monitoring dated 2026-02-24), 15 audit documents, and source-provided contract addresses.
  • Source type: the documentation is Valantis's own site and its publisher is not verified (source reliability: unverified publisher), so mechanics, controls and addresses rest on unconfirmed statements. No independent articles or Blockworks reports are in the workspace.
  • Missing, on-chain: no on-chain read of stHYPE, Overseer, wstHYPE, the multisig, timelock or role holders on HyperEVM. Signers, 4/6 threshold, 48-hour delay and role holders are docs statements only. The measured data covers other Valantis AMM contracts on Ethereum, Base and Arbitrum. A listed Mantle chain read (chain 5000) was not performed.
  • Missing, code: the stHYPE contracts repository is not public, and a bytecode-to-source match is not available for the stHYPE contracts.
  • Missing, numbers: no historical TVL or APY series, no measured fee revenue, no withdrawal-time statistics (the docs promise bi-weekly reports in the future), and no oracle details.
  • Not read: the quarterly stHYPE Health Reports, including the report the docs label the stHYPE External Health Report Audit Q2 2026, are linked as PDFs. A web request for the Q2 2026 report could not retrieve its body, so its author, date, scope and findings are not established here.

102 citations · click markers to view evidence

🔒

Code Security

6/10

Code Security Analysis: Sthype

stHYPE (a liquid staking token on Hyperliquid, about USD 183M TVL) has a solid review record: four independent firms (Three Sigma, Pashov, Guardian, Obsidian) reviewed its staking contracts in seven reports between February 2025 and October 2026. The audited code is in a private repository, so the deployed code cannot be compared with it. Public explorers do show verified source for the stHYPE proxy and its implementation on HyperEVM, but only those two contracts were checked. Bug bounty status is unclear: the protocol's page says none runs, while a bounty platform lists an active program of up to USD 200,000 for the related Valantis STEX product. Confidence in the review depth is high; confidence that the deployed bytecode equals the audited code is low.

Data Completeness

Findings ledger: 15 audit reports, 138 findings in 136 issue groups (two groups merge a table-of-contents entry with its body entry). Read in full: the findings ledger, the audit coverage map, the on-chain control file, the audit index, the governance, transparency and audit-list docs, and the audit digests; the audit sections quoted below were opened to verify specific findings. The remaining audit text reached this report through the ledger and digests.

Gaps:

  • The collected data attributes no deployed address to any audited component. The coverage map has 0 components and 46 audit-named components with no current address, so deployed code was not compared with any audited version.
  • The stHYPE audits review ValantisLabs/sthype-contracts. That repository is private (404 for public readers), so the four audited commits could not be inspected. Code declared these commits as not published.
  • The collected data holds no source for the stHYPE proxies. A separate web check found verified source for the proxy and its implementation on HyperEVM; the Overseer and the staking modules were not checked. The on-chain control read does not include HyperEVM, and Mantle (chain 5000) was not read.
  • Most documentation pages cited here, including the governance, transparency, audit-list and address pages, come from publishers whose status is unverified (source reliability: unverified publisher). Claims that rest on them alone are unconfirmed.
  • Seven audit reports are hosted on raw.githubusercontent.com, which is not verified as auditor-controlled. They count as evidence at lower source confidence.
  • The Statemind Mar 2024 digest was built from the first 80,000 characters only.
  • No code check exists for the unfixed medium/high groups: the repository code was not located. The findings ledger does not carry one Guardian medium finding (Interim Balance Excluded From Backing) as a separate group; it is reported below from the audit text.

Audit Coverage

The audit programme covers the stHYPE staking contracts well. It also covers the older Valantis AMM products, which are separate products and add no review depth for stHYPE staking.

Chains: the protocol's DefiLlama TVL sits on Hyperliquid L1 .

stHYPE staking audits (current product)

The protocol's audit page (source reliability: unverified publisher) lists these under "stHYPE" .

AuditorDateScope and result
Three Sigma (host unverified)2025-02-11stHYPE.sol, wstHYPE.sol, Overseer.sol and others at commit 8e72f83; no critical or high findings
Pashov Audit Group2025-10-13Overseer.sol, HyperCoreUtils.sol, HyperCoreStakingModule.sol; 4 medium and 5 low findings, no critical or high
Pashov Audit Group2025-11-10StakingModuleExternalManagement.sol; 4 low findings
GuardianAudits2025-11-10Staking modules; 1 high (fixed) and 6 informational
GuardianAudits2026-01-05Overseer.sol, stHYPE.sol, wstHYPE.sol, staking module; 1 high (fixed), 2 medium, 4 low, 17 informational
Obsidian Audits (host unverified)2026-04-06StakingModuleExternalAccount is in the stHYPE repository, but the audit summary says it serves Valantis Prime sub-accounts . The same report also reviews StakeAccountManager in the valantis-account repository (Valantis Prime, another product), which is not counted for stHYPE . The report has 1 medium (fixed), 1 low and 3 informational findings across both.
Guardian (host unverified)2026-10-18StakedHypeEscrow.sol; 1 informational (fixed)

No public contest or formal-verification report exists for the stHYPE contracts in the collected data.

Other Valantis products (context only)

These audits cover separate Valantis products: the HOT AMM module, the Core pools, the swap router and Validly. They are listed for completeness and are not counted as review depth for stHYPE staking.

  • Valantis Prime account contracts: Obsidian Audits, 2026-02-21
  • Valantis HOT: WatchPug 2024-03-24 (host unverified), Statemind 2024-04-10 to 2024-04-23, Sherlock public contest 2024-05-15 to 2024-06-04 (host unverified; zero medium or high issues per the report)
  • Valantis Core: Hexens 2024-01, Statemind 2024-02-14 to 2024-03-19
  • Swap router: Statemind, Sep 2024 (host unverified); Validly: Hexens, Sep 2024.
  • The protocol's audit page (source reliability: unverified publisher) also lists Stex audits (Hexens, Zenith, Mar-May 2025) . These belong to another DefiLlama family member (valantis-stex) and are not in this workspace.

Unaudited or unconfirmed areas: the audits do not state a deployed address for any component, and the review of wstHYPE, stHYPE and Overseer at the currently deployed version cannot be linked to deployed bytecode.

Deployed code vs audits

Deployed code was not compared with the audited versions: the coverage map holds 0 deployed components, and 46 audit-named components have no attributed current address. The stHYPE audits name sthype-contracts commits in a private repository, so no equality claim is possible. This is a measurement gap, not a finding that the code differs.

The collected Sourcify rows list factories on Ethereum, Base and Arbitrum with a provider runtime match. Their attribution to Valantis is only a candidate: ownership is not established, so the rows are not used for scoring. Their source files are not materialized, so they could not be compared with any audit. A Mantle factory is listed with only an Etherscan explorer claim (not a runtime match).

The docs (source reliability: unverified publisher) list the stHYPE proxy on HyperEVM . A web check of Sourcify (chain 999) found an exact creation and runtime match for its implementation 0xe71cAF5c1fe56d8897c7B604295D23968049E057, verified 2026-05-23 (source reliability: unverified publisher, but provider data). HyperEVMScan also shows the proxy as source-code verified, an explorer claim . The audited commits are in a private repository, so this verified source could not be compared with any audit.

ComponentDeploymentStatusAudit and revisionCriticalityNote
getTotalBalanceno attributed current addressAudit names it; deployed code not comparedNamed as an affected function in the Pashov Nov 2025 reportcriticalSums staking-module balances for the backing and supply calculation; an audited omission can cause false slashing or block mint and redeem; no deployed address is attributed to it in the collected data
HyperCoreStakingModuleno attributed current addressAudit names it; deployed code not comparedPashov Oct 2025 (stHYPE)criticalDeposits HYPE into HyperCore; an audited failure case makes HYPE drop out of the total balance; no deployed address is attributed to it in the collected data
HyperCoreUtilsno attributed current addressAudit names it; deployed code not comparedPashov Oct 2025 (stHYPE)peripheralNot rated from evidence: the audit text names this file in scope but does not describe its role, so the lowest rating is shown; no deployed address is attributed to it in the collected data
Overseerno attributed current addressAudit names it; deployed code not comparedPashov Oct 2025; Guardian Jan 2026; Three Sigma Feb 2025 (stHYPE)criticalCore stHYPE contract for mint, burn, rebase and slashing; an audited high finding concerned minting after a slash; no deployed address is attributed to it in the collected data
Overseer.Solno attributed current addressAudit names it; deployed code not comparedGuardian Jan 2026 (stHYPE)criticalSame contract as Overseer, named this way in the Guardian Jan 2026 report; no deployed address is attributed to it in the collected data
StakeAccountManagerno attributed current addressAudit names it; deployed code not comparedObsidian Apr 2026 (Valantis Prime repository)highBelongs to Valantis Prime, not to stHYPE staking; it manages staking lifecycle for stHYPE-backed Prime sub-accounts; no deployed address is attributed to it in the collected data
StakingModuleExternalAccountno attributed current addressAudit names it; deployed code not comparedObsidian Apr 2026 (stHYPE)highIn the stHYPE repository but serves Valantis Prime sub-accounts per the audit summary; no deployed address is attributed to it in the collected data
StakingModuleExternalManagementno attributed current addressAudit names it; deployed code not comparedPashov Nov 2025; Guardian Nov 2025 (stHYPE)criticalStaking module whose balance feeds the supply calculation; an audited high finding concerned a reverting burn recipient; no deployed address is attributed to it in the collected data
ConstantSwapFeeModuleno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)peripheralNot rated from evidence: the audit text names this file in scope but does not describe its role, so the lowest rating is shown; no deployed address is attributed to it in the collected data
ConstantSwapFeeModuleFactoryno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)peripheralNot rated from evidence: the audit text names this file in scope but does not describe its role, so the lowest rating is shown; no deployed address is attributed to it in the collected data
DirectSwapno attributed current addressAudit names it; deployed code not comparedStatemind Sep 2024 (Valantis swap router, another Valantis product)highPeriphery swap router that executes swaps through Valantis pools; belongs to a Valantis AMM product, not to stHYPE staking
GaslessSwapno attributed current addressAudit names it; deployed code not comparedStatemind Sep 2024 (Valantis swap router, another Valantis product)highPeriphery swap router that executes swaps through Valantis pools; belongs to a Valantis AMM product, not to stHYPE staking
GMno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)peripheralNot rated from evidence: the audit text names this file in scope but does not describe its role, so the lowest rating is shown; no deployed address is attributed to it in the collected data
HOTno attributed current addressAudit names it; deployed code not comparedStatemind Apr 2024; WatchPug Mar 2024; Sherlock 2024 (Valantis HOT AMM, another Valantis product)highAMM module whose signer is a single point of failure per the audit; belongs to a Valantis AMM product, not to stHYPE staking
ProtocolFactoryno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)highAn audited reorg issue could substitute malicious modules, raising fees or adding malicious ALMs; Valantis AMM product, not stHYPE staking; no deployed address is attributed to it in the collected data
SOTno attributed current addressAudit names it; deployed code not comparedStatemind Apr 2024; WatchPug Mar 2024; Sherlock 2024 (Valantis HOT AMM, another Valantis product)highAMM module whose signer is a single point of failure per the audit; belongs to a Valantis AMM product, not to stHYPE staking
SOTOracleno attributed current addressAudit names it; deployed code not comparedStatemind Apr 2024; WatchPug Mar 2024; Sherlock 2024 (Valantis HOT AMM, another Valantis product)highSupplies the price feed for the AMM mode of HOT; belongs to a Valantis AMM product, not to stHYPE staking
SovereignPoolno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)highPool in which an audit discusses funds stored in the vault; belongs to a Valantis AMM product, not to stHYPE staking
SovereignPoolFactoryno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)highAn audited reorg issue could deposit a victim's funds into a malicious pool; Valantis AMM product, not stHYPE staking; no deployed address is attributed to it in the collected data
UniversalPoolno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)highPool whose swap and flash-loan logic was audited; belongs to a Valantis AMM product, not to stHYPE staking
UniversalPoolFactoryno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)highAn audited reorg issue could deposit a victim's funds into a malicious pool; Valantis AMM product, not stHYPE staking; no deployed address is attributed to it in the collected data
ValantisSwapRouterno attributed current addressAudit names it; deployed code not comparedStatemind Sep 2024 (Valantis swap router, another Valantis product)highPeriphery swap router that executes swaps through Valantis pools; belongs to a Valantis AMM product, not to stHYPE staking
ALMLibno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)peripheralLibrary, struct or test file
AlternatingNonceBitmapno attributed current addressAudit names it; deployed code not comparedStatemind Apr 2024; WatchPug Mar 2024; Sherlock 2024 (Valantis HOT AMM, another Valantis product)peripheralLibrary, struct or test file
Constantsno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)peripheralLibrary, struct or test file
EIP712no attributed current addressAudit names it; deployed code not comparedStatemind Sep 2024 (Valantis swap router, another Valantis product)peripheralLibrary, struct or test file
EnumerableALMMapno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)peripheralLibrary, struct or test file
GaslessSwapIntentHashno attributed current addressAudit names it; deployed code not comparedStatemind Sep 2024 (Valantis swap router, another Valantis product)peripheralLibrary, struct or test file
NonceBitmapno attributed current addressAudit names it; deployed code not comparedStatemind Sep 2024 (Valantis swap router, another Valantis product)peripheralLibrary, struct or test file
PriceTickMathno attributed current addressAudit names it; deployed code not comparedStatemind Mar 2024; Hexens Jan 2024 (Valantis Core AMM, another Valantis product)peripheralLibrary, struct or test file

Remaining 16 components are not shown: further peripheral libraries, structs and constants of the Valantis AMM products, all with no attributed current address. Inventory gaps: no attributed current address for the affected contracts SOT, getTotalBalance, HOT, SOTOracle, SovereignPool, GM and PriceTickMath. The inventory is partial: no first-party complete deployment list is in the collected data.

On-chain control file (factory addresses are candidate attributions only): Ethereum (block 26295639), Base (block 53227151) and Arbitrum One (block 519244140) were read for the Valantis AMM factories. No stHYPE contract on HyperEVM was read, so no pause state, proxy implementation or privilege holder of stHYPE is confirmed on-chain.

Findings by Severity

The ledger holds no critical findings. Seven medium/high groups are not marked fixed, and the Guardian audit adds one more acknowledged medium that the ledger does not list separately. The ledger groups come from Valantis AMM audits or are a remediation note on the stHYPE Overseer, and the added Guardian finding also sits on the Overseer. For each, the repository code was not located, so there is no code check and no effect beyond the audit status alone.

Critical

  • No critical finding in any of the 15 audits.

High

  • The signer acts as a single point of failure (WatchPug on Valantis HOT, 2024-03-24) — High (estimated, an AI estimate; the audit states no rating): the estimate rests on the described drain of up to 43% of pool funds in one block ; title — Audit status: acknowledged — Check: not checked: the audited repository code was not available. This concerns the HOT AMM, a separate product, not stHYPE staking.
  • Fixed high findings, not re-checked: Guardian Nov 2025 "Reverting Burn Recipient Bricks Future Redeems" and Guardian Jan 2026 "Minting Allowed After Slashing" (both fixed per the audits) .

Medium

  • Min Slash Threshold Skews Losses (Guardian remediation review, 2026-01-05; stHYPE Overseer) — Audit status: acknowledged — — Check: not checked: the audited repository code was not available. Small losses below the threshold fall on current holders.
  • Interim Balance Excluded From Backing (Guardian, 2026-01-05; stHYPE Overseer) — Audit status: acknowledged ; the audit digest records that getTotalBalance omits the deprecated interim address, which held about 10.999535619 HYPE on the forked mainnet test, which can cause false slashing or block mint and redeem flows — Check: not checked: the audited repository code was not available.
  • HOT does not handle rebasing scenarios (Statemind Apr 2024) — Audit status: partially fixed per the ledger — Check: not checked: the audited file was not found.
  • ALM doesn't work in AMM mode in the lack of oracles (Statemind Apr 2024) — Audit status: partially fixed — Check: not checked: the audited file was not found.
  • Complete or temporary shutdown of ammSwap (Statemind Apr 2024) — Audit status: acknowledged — Check: not checked: the audited file was not found.
  • Loss of pool manager's commissions (Statemind Valantis Core, Mar 2024) — Audit status: acknowledged — Check: not checked: the audited file was not found.
  • Rounding in favor of the protocol account for the last token in a quote by incorrect price (Statemind Valantis Core, Mar 2024) — Audit status: acknowledged — Check: not checked: the audited file was not found.
  • The other 17 medium groups in the ledger are marked fixed and are not re-checked.

Low / Informational

Low: 18 groups (11 fixed, 6 acknowledged, 1 not stated). Acknowledged lows include "Deposit blockage by liability increase may leave dust" (Pashov Oct 2025), "Rebase does not have protection for sudden balance drops" (Pashov Nov 2025), "Insolvency Can Revert Backing Math" (Guardian Jan 2026) and "Missing delegation lock period check allows silent undelegation failure" (Obsidian Apr 2026).

Informational: 92 groups (63 fixed, 24 acknowledged, 2 partially fixed, 3 not stated). Acknowledged items from Guardian Jan 2026 include "Non-atomic Upgrade Leaves V3 Uninitialized" and "Frontrunning V3 Upgrade". The three Three Sigma informational items carry AI-estimated severity because the audit gives no rating.

Recurring Patterns

  • Rebase and slash accounting: the stHYPE audits repeatedly find edge cases in how supply, burns and backing react to rebases and slashing. Examples: Pashov Oct 2025 "rebase() can be blocked by large supply increase", Guardian Jan 2026 "Extreme Slash Can Freeze Rebase Flow", "Burns Can Revert During Active Rebase" and Min Slash Threshold. All of the high findings in this family were fixed; several informational ones are acknowledged.
  • Rebasing tokens in AMM modules: Statemind's HOT audit reports "HOT does not handle rebasing scenarios" (partially fixed) and a negative-rebase revert in SOT.getReservesAtPrice() (partially fixed).
  • Privileged control: the protocol documents a 48-hour proxy-upgrade timelock (source reliability: unverified publisher) and a 4/6 multisig (source reliability: unverified publisher) . The same registry (source reliability: unverified publisher) says that default-admin transfer delay is zero and no on-chain veto exists . These governance facts are assessed in the governance report.

Code Quality Assessment

  • Review depth is good for the current stHYPE version: seven reports from four firms (Three Sigma, Pashov, Guardian, Obsidian) over 20 months, with remediation reviews. The newest report is dated 2026-10-18.
  • Fix rate is high: all critical-to-high stHYPE findings are fixed per the audits; the unfixed medium items are acknowledged or partially fixed; two sit on the stHYPE Overseer (backing and slash accounting) and the rest belong to the AMM products.
  • Assurance cannot be graded from the available evidence. The stHYPE docs state that no bug bounty runs (source reliability: unverified publisher) . A bounty platform lists an active program for the related Valantis STEX product, with critical payouts up to USD 200,000 (source reliability: unverified publisher); its scope names stHYPEWithdrawalModule but not the staking contracts . Which statement is current is not established. Verified source was confirmed only for the stHYPE proxy and its implementation; the Overseer and the staking modules were not checked.
  • Assessment: the main open question is whether the deployed stHYPE bytecode matches the audited commits. The governance registry (source reliability: unverified publisher) documents an upgrade of all proxy contracts on 2026-04-10 , but the workspace cannot link that deployment to an audited commit. Comparing the verified HyperEVM source with the audited sthype-contracts commits would settle it, but that repository is private.

Score: 6.0/10

Rubric security-score-v2.5; data completeness 0.85

Code security factors, points out of 10 (points)
Audit coverage
5.50 points
Audit depth
7.50 points
Unresolved findings
8.91 points
Assurance
3.50 points
Audit findings not marked fixed (findings)
High · acknowledged
1 findings
Medium · partially fixed
2 findings
Medium · acknowledged
4 findings
FactorLevelEvidenceWeightPoints
Audit coverage (computed)L2supported0.355.50
Audit depthL3supported0.207.50
Unresolved findings (computed)L4supported0.308.91
Assurance—not disclosed0.153.50

37 citations · click markers to view evidence

💰

Economic Security

5.5/10

Economic Security Analysis: Sthype

stHYPE is a liquid staking token for HYPE on Hyperliquid. Users deposit HYPE, receive stHYPE, and the protocol stakes the HYPE on HyperCore (Hyperliquid's native staking layer). The main economic risks are slow exits when large amounts leave at once, no documented guaranteed insolvency backstop (and no qualifying loss-cover fund found), reported high holder concentration and thin DEX depth against large lending collateral; the last two rest on unverified third-party data. Yield comes from HyperCore staking rewards, and restaking adds deployer-paid rewards; the share of own-token incentives in paid yield is not established. The evidence is mostly the team's own docs (unverified publisher) plus several audits, so confidence is moderate. Only the protocol's HyperEVM deployment holds TVL, and no on-chain role data was captured for it.

Scope: the DefiLlama record covers stHYPE on Hyperliquid L1 only. The related Valantis products (Prime trading accounts, the HOT/Sovereign AMM on Ethereum and other chains, STEX) are context, not scored here, except where they hold or route stHYPE funds.

TVL & Liquidity Profile

TVL is about USD 183M, all on one chain, and DEX depth is not measured by code; the available figures are weak and from May 2026.

  • DefiLlama snapshot (collected 2026-10-29): . Category is Liquid Staking and there is no other active chain.
  • Chain profile: Hyperliquid L1 is classed as other, with no L2BEAT stage and no exit window measured . The chain's launch date is not cited in the workspace, so its age is not assessed.
  • Measured DEX depth: none. The depth tool only probes Ethereum tokens and found no candidates . stHYPE and wstHYPE live on HyperEVM, so the 2% and 0.5% depth bands are gaps, not zero.
  • Third-party depth (Yearn risk report, May 2026, source reliability: unverified publisher): about USD 6.26M of stHYPE pools and USD 690K of wstHYPE pools . That is five months old and no current figure was found.
  • Holder concentration (HyperEVMScan, fetched 2026-10-29, source reliability: unverified publisher): the top 5 stHYPE holders own 79.00% and holders 6-10 own 11.70%, so the top 10 own about 90.7% . For wstHYPE the top 10 own 61.53% . The explorer does not label the holders, so some may be protocol or lending contracts and the true holder concentration is unconfirmed. A large holder exit could exceed the third-party May 2026 depth figures above, but no current depth was measured.
  • Exit path (native): withdrawals have no fee and target about 7 days in normal conditions (source reliability: unverified publisher). The docs set a target of about 10,000–50,000 HYPE for the instant-redemption queue; the actual queue size was not measured (source reliability: unverified publisher).
  • Points: Valantis points reward holding stHYPE or wstHYPE (source reliability: unverified publisher). The share of TVL tied to the campaign is not cited, so the campaign effect on TVL is a gap.

Oracle Risk

stHYPE has no market price feed for minting or redeeming; it mints 1:1 and moves its exchange rate through a keeper-triggered rebase computed from on-chain HyperCore balances; Pashov's audit shows two bounds in the rebase code, and the docs' further claims are unconfirmed.

  • Mint and burn do not read a DEX price. Guardian's January 2026 review notes that "stHYPE mints 1:1" .
  • Rate source: Overseer.rebase() computes the new supply from HyperCore precompile balances . A keeper holding REBASER_ROLE triggers it, so the keeper decides timing but not the number.
  • Rebase bounds in code: Pashov's review (October 2025) shows rebase() reverting with AprTooHigh and SupplyDecreaseTooHigh . Current deployed values were not read on-chain. The docs state that "Rebase safety bounds can revert unsafe supply changes (AprTooHigh, SupplyDecreaseTooHigh)." (source reliability: unverified publisher). The docs also report that pending-slash checks can block mint and redeem; this is unconfirmed outside the docs (source reliability: unverified publisher).
  • Past defects, all marked fixed: precompiles return start-of-block state, which could miscalculate supply . A donation could block rebases through the APR cap (Pashov, October 2025, medium, fixed).
  • Downstream: the Prime docs report that HyperLend values wstHYPE with redemption value rather than a DEX price, so a market depeg alone would not liquidate borrowers; this is unconfirmed, as no lender source was read (source reliability: unverified publisher). If the report holds, slashing or a stHYPE contract bug could still lower collateral value.
  • Gaps: no oracle-caused loss is cited. The rebase key-holder identities are on HyperEVM and were not read on-chain.

Flash Loan & MEV Exposure

No flash-loan exploit of stHYPE is documented; the MEV-type issues found were front-running around slashing and share price, and all were fixed.

  • Guardian (January 2026) rated "Minting Allowed After Slashing" high: in the audit's scenario, a minter plugs a pending-slash shortfall with its own deposit and mints 1:1, so the minter's new tokens are slashed later and the loss shifts from existing holders to the minter, bypassing the intended no-mint-or-redeem-until-slash-is-applied policy . Its status is resolved.
  • An audit dated 11 February 2025 (author not named in the excerpt, published on an unverified host) found a medium share-price inflation attack via minting to the zero address, fixed . An unresolved medium remains: Guardian's "Min Slash Threshold Skews Losses" was acknowledged in the remediation review, not fixed, so small slashes land on current holders .
  • A second Guardian review (November 2025) found the high "Reverting Burn Recipient Bricks Future Redeems", resolved .
  • Liquidation MEV on borrowers is a lending-market matter (HyperLend and others), outside this record. The Sovereign pool flashLoan in the HOT audit belongs to another Valantis product.
  • No public-mempool MEV measurement is available. HyperEVM sequencing is not analysed in the workspace.

Tokenomics Risk: Yield, Leverage & Reflexivity

Yield is Hyperliquid staking rewards plus optional restaking rewards; no own-token emission is cited, but the emission share of paid yield, the protocol fee and the revenue split are not established, and optional looping adds leverage outside the core token.

  • Revenue source: stHYPE is "backed by HYPE staked on HyperCore, allowing holders to earn staking rewards" . The docs add that "Every account can only stake to HyperCore. None of them lend or rehypothecate." (source reliability: unverified publisher).
  • Amounts (team figures, not independent): the app shows 1.97% APY , and Prime docs use 2.4% "normal stHYPE yield" as of 15 July 2026 (source reliability: unverified publisher). These are targets or snapshots. The protocol fee rate on rewards was not found; a fee recipient role exists (claimFee).
  • Restaking adds extra reward streams paid by deployers: "they can be HYPE premiums, token incentives, points, or revenue share" (source reliability: unverified publisher). The page lists 10 HYPE per day for the active USDe quote-asset module; the other streams are not live. Surplus yield goes to holders by rebase or is retained at Valantis' discretion.
  • Own-token share: stHYPE is the deposit receipt, and no separate governance or emission token is cited. Points are not a token. The own-token emission share of paid yield is not established: the deployer-reward page lists token incentives among possible streams, and no source quantifies them.
  • Leverage in the core position: none is reported, and this is unconfirmed outside one third-party review. Staking Rewards' review states "Yield was sourced from native HYPE staking with no leverage." (source reliability: unverified publisher).
  • Optional leverage via Prime/HyperLend (reported by the Prime docs, unconfirmed against HyperLend itself): the docs say HyperLend allows "59% maximum LTV, 74% liquidation threshold, and 15% liquidation penalty" for USDC borrowing against wstHYPE (source reliability: unverified publisher). The looped strategy runs at up to a 77.77% e-mode LTV (source reliability: unverified publisher), with unwind by liquidation "below health factor 1" (source reliability: unverified publisher). At maximum e-mode leverage the slashing buffer is materially smaller.
  • Reflexive paths: slashed HYPE is burned, not repaid to holders (source reliability: unverified publisher), and slashing socializes through rebase. No own-token spiral is cited, and reflexive exposure through restaking reward tokens is not established; a slash plus liquidations of looped wstHYPE could compound.

Solvency, Liquidation & Bad Debt

Backing is HYPE staked on HyperCore, which is checked on-chain by the rebase logic, but exits can take 30 to 183 days for the part held with deployers, and no guaranteed insolvency backstop is documented.

  • Backing against claims: stHYPE is minted 1:1 against HYPE, and mint, burn and redeem are checked against a pending-slash state (see Oracle Risk). Backing vs. liabilities was not independently attested in the workspace; stake-account balances were not read on-chain.
  • Exit path: users call burn, burnAndRedeemIfPossible or redeem. The integration docs say burnAndRedeemIfPossible can redeem instantly only up to maxRedeemable(), and the remainder enters a queue (source reliability: unverified publisher).
  • Maximum delay: deployer stake has "documented recall periods range from 30 to 183 days" (source reliability: unverified publisher). The docs also state that "Prolonged withdrawals remain possible if requests exceed what the standard accounts and new deposits can absorb." (source reliability: unverified publisher). Staking Rewards puts about 15% of TVL in modules with lockups of up to 90 days (source reliability: unverified publisher); that differs from the 183-day figure, and the current share is not confirmed. The only active deployer account is the USDe quote-asset module (source reliability: unverified publisher).
  • Pause power: the docs state that burns and redemptions can be paused and that the pause is bounded on-chain; this bound is not confirmed in audits or on-chain reads: "Pause duration is bounded on-chain (> 0 and <= 14 days)." (source reliability: unverified publisher). Off-chain alerts do not auto-submit pauses.
  • Restaking escrow pause (separate from the core redemption pause, and relevant only to restaked positions): the integration docs report that setPaused blocks deposits, withdrawal requests, cancellations and execution together, with "no automatic expiry or maximum pause duration" (source reliability: unverified publisher). A keeper holds this pause role in the same docs. This is an unconfirmed report; the escrow audit in the available sources does not mention the pause.
  • Loss handling: slashes reduce supply for all holders via rebase. No qualifying loss-cover fund was found. The docs state "There is currently no guaranteed insolvency backstop for user losses." (source reliability: unverified publisher).
  • Open audit item: Guardian's "Insolvency Can Revert Backing Math" (low) and "Interim Balance Excluded From Backing" (medium) were acknowledged, not fixed .
  • Debt positions: stHYPE itself has none. Liquidation happens in third-party lenders. Yearn's May 2026 report counts about USD 65M of wstHYPE used as collateral against about USD 690K of wstHYPE DEX liquidity (source reliability: unverified publisher). The liquidation access rules and bad-debt records of those lenders are not in the workspace.
  • Bad-debt record of stHYPE: none found. No exploit or bad debt for stHYPE appears in the DefiLlama hacks file, which is empty . The absence of a record does not prove none occurred.

Incentive Alignment

Incentives rest on staking yield, with the emission share of paid yield not established, and no guaranteed insolvency backstop is documented; the search found no qualifying loss-cover fund, so who absorbs a loss beyond slashing's rebase effect is not established.

  • Mercenary capital: the points program rewards holding and DeFi use of stHYPE/wstHYPE, and 260M weekly points were fully distributed by 4 June 2026 (source reliability: unverified publisher). Whether TVL leaves when points stop is not measured.
  • Restaking: restakers accept a 14-day withdrawal cooldown for extra rewards, and the reported escrow pause has no stated maximum duration (see Solvency) (source reliability: unverified publisher). The cooldown adds to the ordinary exit process, and restakers share the delegation-related withdrawal constraints with ordinary stHYPE holders, so ordinary holders can still face prolonged exits from the same delegation commitments (source reliability: unverified publisher).
  • Loss cover: no qualifying loss-cover fund was found in the available sources. Staking Rewards reports the treasury is under 0.25% of TVL with no formal activation rules (source reliability: unverified publisher). A treasury without a stated purpose is not a fund. No bug bounty is active (source reliability: unverified publisher).
  • Past payouts: no incident with user loss is recorded, so there are no payout rows.
  • Governance of the protocol sits with a multisig documented as 4/6 (source reliability: unverified publisher) and a 48-hour proxy-upgrade timelock (source reliability: unverified publisher). Signers and threshold are documentation values (updated 2026-04-07), not confirmed on-chain.

Economic Attack Vectors

The credible vectors are slashing losses, a run on the exit queue, and liquidation cascades in lending markets that hold wstHYPE.

  • Slashing: the deployer-stake page describes HIP-3 deployers (builder-deployed perpetual markets), where validator-elected slashing addresses irregular deployer inputs and penalties can reach the entire stake, which remains slashable during unstaking (source reliability: unverified publisher). The page also covers HIP-4 outcome markets. The sources do not establish which of these rules apply to the active USDe quote-asset module, so that module's slashing exposure stays uncertain. Pashov's review noted that base-staking slashing was not yet enabled on Hyperliquid at that time .
  • Exit run: if a large share of holders burn at once, the instant queue (target 10,000–50,000 HYPE, size not measured) empties, then fast-withdrawal accounts, then deployer recalls. The secondary market would then price the delay. The explorer's top-10 figure (about 90.7%, unlabelled holders, unconfirmed) suggests a large-holder exit is possible.
  • Liquidation cascade: the Yearn report (May 2026, source reliability: unverified publisher) counts about USD 65M of wstHYPE collateral against about USD 690K of wstHYPE DEX liquidity, and this is unconfirmed by current data. The Prime docs report that HyperLend's oracle follows redemption value, which would prevent depeg liquidations but not slashing-driven ones; this is unconfirmed. If the figures still hold, impact on borrowers would be high; the likelihood cannot be banded without current depth and lender data.
  • Price manipulation: no spot price feeds mint or redeem, so manipulation of a DEX pool does not change protocol accounting. It can affect lenders that use DEX prices.
  • Dual-token interface: stHYPE and wstHYPE share state, and the docs warn of a "significant vulnerability for certain DeFi integrations" (source reliability: unverified publisher).
  • Bridge risk: none documented. The docs call CoreWriter a native precompile, not a bridge, and Staking Rewards accepted that view.
  • Dependencies (each is critical to principal): Hyperliquid HyperCore staking and validator set (base chain); HyperCore deployer stake rules (HIP-3/HIP-4 and quote assets, whose application to the active module is not established); the USDe quote-asset module (Guardian's November 2025 review notes that the staking module forwards deposits to a generic stake account and trusts it to whitelist a quote asset; the review does not give the module's launch date or establish its failure impact ); Prime sub-accounts that move stHYPE-owned stake (Obsidian audits of April 2026 and February 2026 cover these). HyperLend, Felix and others are downstream, not under the deposit.
  • Dependency graph: the only edge is LIVES_ON Hyperliquid L1 at share 1.0, with no protocol edges, no oracle providers and composition status missing .
  • On-chain control file: it reads the Ethereum HOT AMM pool of another Valantis product (a 2-of-3 Safe and a paused pool manager at block 26295639), not stHYPE. Hyperliquid contracts were not read, so stHYPE roles stay unconfirmed on-chain .

Aggregator Scores

No curated security score is available in the workspace for this protocol, so these checks rely on direct evidence.

  • DeFiLlama: TVL USD 183.2M on Hyperliquid L1 only, category Liquid Staking, captured 2026-10-29 .
  • CertiK, DeFiSafety and Exponential files are absent from the workspace. A third-party rating (Staking Rewards, 25 June 2026, source reliability: unverified publisher) moved stHYPE from CCC+ to BB+ after the team published a multisig, a timelock and monitoring docs (source reliability: unverified publisher). It also lists no bug bounty, no live token governance and a thin treasury as open gaps. It is a commissioned process with the team, so it supports but does not set any level.
  • Comparison: the rating's remaining gaps (no bug bounty, a thin treasury, lockups, thin liquidity) match what the detailed read finds on exit delay and loss cover.

Data Completeness

Data on yield mechanics, rebase bounds, audits and governance is good; independent on-chain measurement of backing, roles and DEX depth is missing.

  • Available: DefiLlama TVL, eight stHYPE-relevant audit reports (Pashov Oct and Nov 2025, Guardian Nov 2025 and Jan 2026, Guardian StakedHypeEscrow Oct 2026, Obsidian Feb and Apr 2026, and an unattributed Feb 2025 report), full docs on withdrawals, roles, monitoring and Prime.
  • Missing: an independent measure of holder concentration (explorer data is from an unverified publisher and unlabelled); on-chain reads of Hyperliquid contracts (stake-account balances, role holders, multisig signers); current HyperEVM DEX depth; the protocol fee rate; the Health Reports; holders labelled by type; the effective share of TVL in deployer lockups; bad-debt records of the lending markets that take wstHYPE.
  • Provenance: several audits are hosted on an unverified host (raw.githubusercontent.com), and the docs are tagged as an unverified publisher, so levels rest on audit text where possible and confidence is lowered, not the protocol penalised.
  • Dated inputs: the Yearn report is from May 2026 and the Staking Rewards rating is from June 2026; both predate the October 2026 snapshot.
  • Confidence: moderate (about 0.55).

Score: 5.5/10

Rubric security-score-v2.5; data completeness 0.75

Economic security factors, points out of 10 (points)
Solvency redemption
5.50 points
Oracle design
7.50 points
Dependency risk
3.50 points
Liquidity concentration
3.50 points
Yield design
5.50 points
FactorLevelEvidenceWeightPoints
Solvency redemptionL2supported0.355.50
Oracle designL3supported0.257.50
Dependency risk—not disclosed0.153.50
Liquidity concentration—not disclosed0.103.50
Yield designL2supported0.155.50

61 citations · click markers to view evidence

👥

Team & Transparency

3.5/10

Social & Team Transparency: Sthype

stHYPE is run by Valantis Labs. Unverified publishers name two co-founders and a Panama-registered operating company, but no official team page or qualifying source confirms them, so those names stay unconfirmed. What is verifiable is the project's public work since early 2024 under the Valantis Labs name. No hack or exploit is recorded. Public development is steady but thin, and the docs state, unconfirmed, that no bug bounty runs for stHYPE. Confidence is low for identity and for track record.

Team Identity

Named people and the operating entity appear only in unverified publications and no source verifies them, so they are unconfirmed. The project's public work under the Valantis Labs name is verifiable, but it is not tied to an accountable person.

  • Verifiable public work: An audit of Valantis HOT carries a delivery date of Mar 24, 2024 . The Valantis factory contracts on Ethereum were created on 2024-08-23 . The official GitHub organisation lists public repositories, including the Hybrid Order Type AMM and core repositories . No official page names an accountable person, and the official GitHub organisation page says "This organization has no public members."

  • Legal entity: The Terms of Service say the site is "operated by Valantis Panama Inc., a Corporation registered under the Laws of the Republic of Panama" (source reliability: unverified publisher). A Crunchbase profile lists the same legal name, with "Paragon Labs" as an alternative name (source reliability: unverified publisher). A third-party page says a separate UK entity was dissolved in October 2024. That claim is unconfirmed, and no source links it to the Panama operator.

  • Co-founders: Greg Vardy is listed as co-founder (unconfirmed), previously Chief Technology Officer at Nethermind (source reliability: unverified publisher). Deven Matthews is listed as co-founder, with a prior role at Nethermind according to a startup directory (source reliability: unverified publisher). The latest commit to the official audits repository was authored by a GitHub account named devenmatthews . The sources do not state that this account belongs to the co-founder, so it is not treated as identity evidence.

  • Other names: A professional-network page lists Eduardo Carvalho as Co-Founder and CTO, and Addison Spiegel as an advisor (source reliability: unverified publisher). CoinDesk (August 2025) reports that stHYPE came from the Thunderhead team: "Addison Spiegel, founder of Thunderhead, the team behind stHYPE, will serve as an advisor to Valantis." (source reliability: unverified publisher).

  • Team size: A Staking Rewards case study says the team is small: "Six full-time people means key-person dependency." (source reliability: unverified publisher).

  • Conflicts: Greg Vardy's current title (CEO or co-founder) differs between profile sites. Eduardo Carvalho's prior work is not verified.

Track Record

No incident is recorded against this protocol or its family, but the evidence on stress behaviour is weak.

  • Operating history: The family was listed on 2025-04-03, per workspace metadata. CoinDesk reports that Valantis acquired stHYPE in August 2025, when stHYPE held about $180 million in TVL (source reliability: unverified publisher). The current DefiLlama TVL is about $183 million, all on Hyperliquid L1 .
  • Incidents: The DefiLlama hacks snapshot holds no records for this protocol or its family . The Staking Rewards study says "The exploit history was clean." (source reliability: unverified publisher). No incident and no post-mortem were found in the reviewed sources.
  • Delivery on promises: After that external rating, the same study describes a team that shipped changes. The study claims, unconfirmed, that on April 7, 2026 the multisig threshold rose from 3/5 to 4/6 with all six signers disclosed, and that proxies moved behind a 48-hour timelock on-chain (source reliability: unverified publisher). The protocol's own governance page also states "stHYPE proxy upgrades are subject to a 48-hour on-chain timelock." (source reliability: unverified publisher). The same page lists a holder veto and token governance as roadmap direction, and says there is "currently no active on-chain veto mechanism enforcing upgrades." (source reliability: unverified publisher). This is unconfirmed by a qualifying source, so delivery of that roadmap item is not established.
  • Disclosure history (earlier gaps): The Staking Rewards study says a February 2026 rating found the multisig threshold, signer count and identities undisclosed, and "An admin key transfer event on-chain with no corresponding documentation" (source reliability: unverified publisher). The study says these gaps were later closed, but this rests on one weak-tier source and stays unconfirmed.
  • Stress history: Docs say quarterly health reports cover key operations. No market-wide stress event or outage is cited in a non-weak source, so resilience under stress is not established here.
  • Own tokens: stHYPE and wstHYPE live on HyperEVM. The Ethereum token check has no candidates , so no depeg history was measured. Mainnet proxies are listed in the docs (source reliability: unverified publisher).
  • Prior projects: The sources name Nethermind as a prior employer of Greg Vardy and, per a directory page, of Deven Matthews. They give no prior-protocol track record.

Social Presence

Public development is active at a low rate, and community metrics were not captured.

  • GitHub: The official audits repository has commits in October 2026, July 2026, April 2026 (two), March 2026 and November 2025 . That is five distinct months in the last twelve. These commits mostly upload audit PDFs, not code. The valantis-stex repository last changed on 26 March 2026 . Other repositories on the organisation page show no 2026 updates.
  • Twitter/X: The official site footer links the @ValantisLabs account. Follower counts and engagement were not captured.
  • Discord and blog: The site links a Discord server and a Mirror blog. Member counts, post cadence and discussion quality were not captured.
  • Docs: The governance page is dated 2026-10-16 (source reliability: unverified publisher). The monitoring page is dated 2026-02-24.

Transparency Indicators

Documentation is detailed and audits are public, but the docs state that no active public bug bounty runs for stHYPE and that no guaranteed insolvency backstop exists. These doc statements are unconfirmed by a qualifying source.

  • Addresses and roles: The docs list mainnet stHYPE, OverseerV1 and wstHYPE proxies, and a separate Roles Registry. The address page is dated about eight months before this analysis (source reliability: unverified publisher).
  • Audits: The docs list seven stHYPE audits plus core, periphery and exchange audits, including (source reliability: unverified publisher). This analysis did not check that every listed report is published.
  • Reports: The latest dated health report listed is April 1 2026, over 180 days before this analysis, so it is historical context and not a current-state snapshot. A "Q2 2026" external health report audit is also listed without a date. "stHYPE Health Reports are internal documents published roughly once a quarter." (source reliability: unverified publisher). The Staking Rewards study says treasury is under 0.25% of TVL with no formal activation rules (source reliability: unverified publisher).
  • Bug bounty: The docs state "stHYPE currently does not run an active public bug bounty program." (source reliability: unverified publisher). A Remedy program exists for the Stake Exchange AMM (another product, with a withdrawal module for stHYPE) with a $200,000 maximum critical payout (source reliability: unverified publisher). That listing is marked Active and names STEXAMM and stHYPEWithdrawalModule as in scope (source reliability: unverified publisher). The two statements have different scopes: the docs describe stHYPE, and the Remedy listing covers the exchange AMM and its stHYPE withdrawal module. This analysis does not treat them as contradictory.
  • Loss cover: The docs state "There is currently no guaranteed insolvency backstop for user losses." (source reliability: unverified publisher).
  • Open source: The core Valantis and stHYPE exchange repositories are public. Whether the stHYPE staking contracts have public source was not established here.

Red Flags

  • No official team page; the GitHub organisation shows no public members, and named people and the operating entity rest on unverified publications that no source verifies.
  • Small team of six, with key-person dependency per the Staking Rewards study.
  • The docs state no active public bug bounty for stHYPE and no guaranteed insolvency backstop (both unconfirmed by a qualifying source). A separate listing covers the exchange modules.
  • The docs and the Staking Rewards study say token voting and a holder veto are not yet live (source reliability: unverified publisher). This is unconfirmed by a qualifying source.

Positive Signals

  • Public work under the Valantis Labs name since early 2024, and a Panama entity and two co-founders claimed by unverified publications.
  • No recorded hack or exploit in the hacks snapshot.
  • Many published audits, and health reports listed through April 2026 (historical).
  • An upgrade timelock and a disclosed signer set after an external review, claimed by the Staking Rewards study and the protocol's own docs (both unconfirmed by a qualifying source).

Data Completeness

  • Available: docs, Terms of Service, audit lists, GitHub commit pages, third-party profile pages and one news article.
  • Missing: follower and member counts for Twitter, Discord and Telegram, blog cadence, and an official team page.
  • No articles from the collected-article set were available. The Staking Rewards study and CoinDesk article are weak-tier external publications, so their claims remain unconfirmed.
  • This analysis rests on collected documents and not on live social-media scraping.

Score: 3.5/10

Rubric security-score-v2.5; data completeness 1.00

Team transparency factors, points out of 10 (points)
Team identity
1.00 points
Disclosure
5.50 points
Development activity
5.50 points
FactorLevelEvidenceWeightPoints
Team identityL0supported0.451.00
DisclosureL2supported0.355.50
Development activityL2supported0.205.50

Track record: 6.1/10

Rubric security-score-v2.5; data completeness 0.80

Track record factors, points out of 10 (points)
Operating history
5.50 points
Incident record
7.50 points
Resilience
3.50 points
FactorLevelEvidenceWeightPoints
Operating history (computed)L2supported0.305.50
Incident record (computed)L3supported0.507.50
Resilience—not disclosed0.203.50

34 citations · click markers to view evidence

🗳

Governance

4.7/10

Governance Analysis: Sthype

stHYPE (Valantis' liquid staking token for HYPE on HyperEVM) is run by a 4-of-6 multisig, not by token voting. Proxy upgrades wait 48 hours, but the official docs list no delay for parameter changes or role grants, and state that the two-day admin transfer delay does not apply to restaking setters or operational-role changes. Signer identities are not disclosed. A user should treat this as a team-controlled protocol with partial safeguards. Confidence is moderate: the key facts come from the protocol's own docs, and the on-chain data supplied does not cover HyperEVM.

Governance Structure

The protocol is a hybrid of a multisig and role-admin keys. No token vote controls changes today.

  • Type: multisig/role-admin. The docs say .
  • Voting mechanism: none active. The docs state . A holder-veto design modelled on Lido's Dual-Governance Framework is planned, with timing tied to "Valantis Governance" adoption in 2026 and no committed dates .
  • Proposal process and quorum: none, because there is no voting. The vote-concentration dataset lists no venue (no Snapshot space or Governor), so no top-10 voter share or quorum could be measured.
  • Timelock: 48 hours, for proxy upgrades only .
  • Emergency bypass: the docs list pause roles on stHYPE (PAUSER_ROLE: pauseTransfer, pauseMint, pauseBurn, pauseRebase) but give no holder address . The Overseer admin role also holds pause controls. The restaking escrow StakedHypeEscrow (SHE) has a separate keeper that can pause deposits, withdrawal requests and execution . The core safety matrix also lists operator-confirmed pauseRedemption and pauseBurns .
  • No active veto:

Admin Key Analysis

One 4-of-6 Safe multisig holds role administration for stHYPE, wstHYPE and Overseer, and owns the proxy admins. Signer identities are not published.

  • Controller: the registry lists 0x97dEe0eA4Ca10560f260A0f6f45bdC128A1D51f9 as DEFAULT_ADMIN_ROLE for stHYPE, Overseer and wstHYPE, and as the multisig-administered owner of the ProxyAdmin contracts, with upgrades subject to a 48-hour timelock . The threshold is 4/6, raised from 3/5 on 2026-04-07 .
  • Signers: six addresses are listed with no names or affiliations in the registry. A community-written risk review (source reliability: unverified publisher) says the same . Whether any signer sits outside the team is not confirmed.
  • Powers: DEFAULT_ADMIN_ROLE administers roles, staking-module and parameter controls and pause controls on Overseer. The registry lists MINTER_ROLE (mint), BURNER_ROLE (burn) and REBASER_ROLE (syncSupply) on stHYPE, so role administration reaches those functions .
  • Operational roles: MANAGER_ROLE, REBASER_ROLE and FEE_RECIPIENT_ROLE on Overseer sit at separate addresses. The docs describe keepers as permissioned but unable to custody funds .
  • Admin power removed: on 2026-04-10 the team removed the admin's ability to block individual addresses from transferring stHYPE and wstHYPE .
  • Benefit: a multisig and keepers can pause or fix problems quickly. The registry says address-level transfer restrictions are separate from burn and claim accounting . That statement concerns address-level restrictions only. Burns and redemptions can still be paused by the admin for a bounded period, and withdrawals can take much longer than the upgrade delay .
  • Restaking escrow (SHE): a keeper holds MANAGER_ROLE (setPaused, rescueTokens limited to unrelated ERC-20 tokens). The multisig holds DEFAULT_ADMIN_ROLE with setCooldown (1 second to 180 days) and setRewardDuration (1 second to 7 days) . The docs checked on 2026-10-18 give a 14-day cooldown . The escrow pause has no automatic expiry or maximum duration, so restakers cannot exit while it lasts.
  • On-chain confirmation: the supplied on-chain control data covers Ethereum, Base and Arbitrum and does not include HyperEVM. The available on-chain read tool queries Ethereum only; a read of the multisig address there found no deployed code at block 26295910. No HyperEVM read was possible, so the 4-of-6 threshold, signer set and role holders are documented but not confirmed on-chain.
  • Other Valantis deployments (context): on Ethereum, at block 26295639, the ProtocolFactory deployer is a Safe with . These factories belong to the Valantis AMM framework, not to the stHYPE token contracts. They are context only.

Upgradability

The stHYPE-related contracts described in the docs are upgradeable proxies, and the official docs state a 48-hour delay on their upgrades.

  • Pattern: proxies with separate ProxyAdmin contracts. Docs describe stHYPE, Overseer, wstHYPE and seven staking modules, each with a ProxyAdmin .
  • Who upgrades: the multisig, through the ProxyAdmin owner path. On 2026-04-10 the team routed all of them behind the 48-hour timelock .
  • Timelock contract: for the restaking escrow, the official docs name a protocol timelock (0xc5DEd4F7F53919c714059a17d31371aE847E23D2) as ProxyAdmin owner, with the multisig as proposer, canceller and executor . A community review (source reliability: unverified publisher) says the same timelock owns all three core ProxyAdmin contracts . The official registry reviewed does not confirm that ownership for the core proxies.
  • Core invariants: an upgrade can replace token and staking logic, including balance logic. Past upgrades changed the Overseer implementation, for example 2026-02-19 for slashing handling.
  • No veto: holders cannot block an upgrade during the 48 hours. Exiting is the only option, and the exit can take longer than the delay.

Parameter Controls

Parameters and roles are changed by the admin multisig. The official docs list no delay for these changes, on the restaking escrow or on core contracts.

  • Bounded: Overseer.setPauseDuration must be above 0 and at most 14 days . It is set to 7 days in V3.
  • Delay on role changes: for the restaking escrow, the official docs say a two-day default-admin transfer delay does not delay setters or operational-role changes . The docs do not say whether any other delay applies to those changes, so that stays unconfirmed. For core contracts the official docs list no delay for role grants or parameter changes. The default-admin transfer delay is a separate control, and the core one is zero . A community review (source reliability: unverified publisher) says role grants and parameter updates run without delay .
  • Example change: the APR threshold for rebases was set to 500 bps on 2025-12-08 .
  • Harm potential: on the restaking escrow, a cooldown change affects future withdrawal requests, and the docs say existing requests keep their claim date. For core contracts, a grant of a minting or rebasing role could take effect without a wait only if no delay applies; no delay is documented and none is confirmed on-chain. Assessment: the main governance gap is that the 48-hour timelock is documented for code upgrades only, and no delay is documented for setters.
  • Deployer allocation: the docs say deployer stake moves "only under timelock" . The curation process for deployer selection is not detailed.

Decentralization Assessment

Control is centralized in one multisig whose signers are not publicly identified. The 48-hour delay and the removed freeze power are real improvements.

  • Spectrum: closer to centralized than to decentralized. Upgrades and role changes depend on four of six unnamed keys. Core pauses sit with the Overseer admin role and the undisclosed PAUSER_ROLE holder, and the restaking escrow keeper can pause that escrow alone.
  • Roadmap: the docs promise a dual-governance veto model tied to Valantis Governance adoption in 2026, without dates. Disclosures on quorum and timelock parameters will follow once governance execution is active.
  • Key-person risk: one signer cannot change roles or upgrade alone, but the restaking escrow keeper can pause that escrow alone. A compromise of four keys could change restaking setters and operational roles with no delay documented and, after 48 hours, upgrade contracts. Whether any delay applies to core role grants is not confirmed.
  • Token distribution: no governance token is wired in, so voting power is not measurable. An explorer snapshot of stHYPE token holders (source reliability: unverified publisher) reports, unconfirmed by another source, . That figure measures token holders, not votes, so it says nothing about governance power today.

Governance Risks

  • Single point of failure: the multisig holds role administration and upgrade proposals and the Overseer admin pause controls; a separate keeper can pause the restaking escrow. Signer key hygiene is the main exposure, along with the keeper key.
  • Changes without a documented delay: the two-day admin transfer delay does not apply to restaking setters or operational-role changes, and no other delay is documented for them. The escrow pause has no maximum duration. Core role administration has no listed delay.
  • Attack vectors: none from voting, since voting is not active. Exposure shifts to multisig signer compromise or collusion.
  • Voter apathy and quorum: insufficient data, because no vote venue exists.
  • Centralization creep: the trend is toward less admin power so far: threshold up from 3 to 4 and the freeze power removed. The registry promises a changelog of future changes.
  • Exit window versus delay: a normal withdrawal targets about 7 days . Deployer recall periods can run 30 to 183 days . The restaking cooldown is 14 days. A 48-hour upgrade delay is shorter than the time a user needs to exit fully.

Data Completeness

Available: protocol governance and roles registry pages (dated 2026-10-16), the restaking escrow and risk-control pages, the contract-address and audit pages, an on-chain snapshot for Ethereum, Base and Arbitrum, and one community risk review.

Missing:

  • On-chain state for HyperEVM (the stHYPE chain): the multisig threshold, signers, timelock contract and role holders are documented only.
  • Signer identities and affiliations.
  • The holder of PAUSER_ROLE.
  • Any voting venue or vote data.
  • The Roles Registry lists the timelock only for proxy upgrades. For core contracts, delays on other changes are not listed. For the restaking escrow's setters, only the exclusion of the two-day transfer delay is documented; whether another delay applies is not stated.
  • A HyperEVM on-chain read: the available tool reads Ethereum only.

Score: 4.7/10

Rubric security-score-v2.5; data completeness 0.60

Governance factors, points out of 10 (points)
Admin control
5.50 points
Change delay
3.50 points
FactorLevelEvidenceWeightPoints
Admin controlL2supported0.455.50
Change delay—not disclosed0.303.50
Decision process—not applicable——

32 citations · click markers to view evidence

🔗

Cross-Analysis

Cross-Risk Analysis: Sthype

stHYPE (a liquid staking token for HYPE on Hyperliquid, about USD 183M TVL, all on Hyperliquid L1) has no recorded hack. Its main risks come from combinations: reported holder concentration and thin DEX depth (both from unverified, partly stale data) meet slow exits, a team-controlled multisig and no documented insolvency backstop. Assessment: the most likely holder loss is a temporary discount or a delayed exit under stress, not a theft. The worst case is a compromise of the admin multisig, which the 48-hour upgrade delay does not cover for role changes. Confidence is moderate to low, because the protocol's controls come from its own docs (source reliability: unverified publisher), and no stHYPE contract on HyperEVM was read on-chain.

Top Risks for a Holder

1. Exit stress: delayed or discounted exits for stHYPE and wstHYPE holders

  • What can go wrong: a large holder exit, or a slash, empties the instant-redemption queue (a target of 10,000–50,000 HYPE). Later exits then depend on fast-withdrawal accounts and deployer recalls of 30 to 183 days, while the DEX market looks thin (a May 2026 third-party figure) and holders look concentrated (unlabelled explorer balances, source reliability: unverified publisher). The docs say . The same report cites top-10 stHYPE holders owning about 90.7% . It also cites DEX liquidity of about USD 6.95M in May 2026 . The protocol docs say operators can also pause burns and redemptions, bounded to 14 days at a time; this bound is not confirmed in audits or on-chain reads .
  • Magnitude: the exit delay applies to all USD 183M. A third-party review puts about 15% of TVL (roughly USD 27M, arithmetic on that figure) in modules with lockups of up to 90 days . The current share is not confirmed, so this is an indicative size, not a ceiling. Holders who must sell on a DEX take the discount. The top-10 balances are about USD 166M at the explorer's valuation, though the holders are unlabelled and may be contracts, so true concentration is unconfirmed. Valantis Prime traders face a separate limit: if the protocol ends a Prime account, balances stay inaccessible for at least seven days . This is a docs statement (source reliability: unverified publisher) and is outside the stHYPE redemption queue.
  • Likelihood: medium for a visible secondary-market discount or queue delay in a stress event; low for a prolonged freeze. No withdrawal-time statistics or depeg history exist, and no stress event is cited . Missing facts that would move the band: current HyperEVM DEX depth, the live instant-queue size, the current deployer-locked share, and holder labels.

2. Admin multisig compromise or misuse (stHYPE, wstHYPE, OverseerV1, restaking escrow)

  • What can go wrong: four of six unnamed signers (or four stolen keys) can change roles and parameters, and can upgrade contracts after 48 hours. The governance report finds no documented delay for core role grants or setters . Holders cannot veto, and a full exit can take longer than the delay . Nothing about the multisig, timelock or roles is confirmed on-chain for HyperEVM.
  • Magnitude: up to the full USD 183M in the worst case, because an upgrade can replace balance logic . Keepers do not custody funds, which limits routine-key damage.
  • Likelihood: low. The protocol docs (source reliability: unverified publisher) state a 4/6 threshold, a 48-hour upgrade delay, and removal of the address-freeze power on 2026-04-10. None of these is confirmed on-chain. There is no incident history. The anonymous team and unconfirmed signer set keep this above "very low". On-chain confirmation of the threshold, signers and timelock would lower the band.

3. Slashing, deployer exposure and socialised losses (stHYPE, wstHYPE, restaked positions)

  • What can go wrong: the only active deployer account is the USDe quote-asset module. Deployer-stake penalties on HIP-3 (builder-deployed perpetual markets) "can reach the entire stake, which remains slashable during the unstaking queue" . Slashed HYPE is burned and the loss is spread across holders by rebase, with no guaranteed backstop . The treasury is under 0.25% of TVL with no activation rules .
  • Magnitude: bounded by the stake held in the affected accounts, which is not measured. The third-party review puts about 15% of TVL (USD 27M) in modules with lockups of up to 90 days; the current share is not confirmed and is not a ceiling for slash exposure. Exposure could be larger if base-staking slashing is switched on. Pashov noted base slashing was not enabled at its October 2025 review. Restaked shares keep slashing exposure during the 14-day cooldown .
  • Likelihood: low. No slashing history is cited. Which HIP-3 or HIP-4 rules apply to the USDe module is not established; that fact and the module's stake size would decide the band.

Combined Attack Vectors

Vector 1: Key compromise inside the exit window

  1. Attackers take four of the six multisig keys, or the signers collude. The signers are not named .
  2. They grant a role or change a parameter. No delay is documented for these steps. Alternatively they queue a malicious upgrade, which waits 48 hours.
  3. Informed holders try to exit, but the normal exit takes about 7 days, the DEX is thin, and burns can be paused by the same admin.
  4. The remaining holders carry the loss. The docs state no guaranteed insolvency backstop. No bug bounty is confirmed for the stHYPE staking contracts: the docs say none runs, while a separate listing covers a related product .
  • Domains: governance (no veto, no role-change delay), social (anonymous team, six-person team), economic (slow exit, no backstop), code (private audited repository, so deployed logic is not comparable).
  • Likelihood: low. The attack needs four keys; the 48-hour upgrade delay and the removal of the freeze power apply. Evidence is incomplete, since no HyperEVM read exists.
  • Impact: catastrophic.

Vector 2: Backing-accounting edge case during a slash, with unverified deployed code

  1. A slash or a balance drop occurs. Guardian's acknowledged medium finding says getTotalBalance omits the deprecated interim address, which can cause false slashing or block mint and redeem .
  2. A second acknowledged medium pushes small slash losses onto current holders .
  3. Rebase can revert on supply-change bounds (AprTooHigh, SupplyDecreaseTooHigh), and the economic report notes a past case where a donation could block rebases (fixed).
  4. Mint and redeem can be blocked by the defect itself, per the audit. The audited commits are in a private repository, so nobody can confirm that the deployed code matches the reviewed version .
  • Domains: code (acknowledged medium findings on Overseer), economic (rebase bounds, no backstop), governance (the admin can pause burns and redemptions, which could lengthen any stall; no source says the pause would be triggered by these defects).
  • Likelihood: low to medium. Rebase and slash edge cases recur across the audits. All high findings of this family were fixed. The interim-address amount found in testing was small (about 11 HYPE).
  • Impact: moderate (frozen mint or redeem, a small mis-allocation). It becomes severe only if combined with a real large slash.

Vector 3: Slash, then liquidation stress in lending markets holding wstHYPE (economic-domain scenario)

  1. A slash lowers the redemption value of wstHYPE (a contract bug is a further, unspecified trigger).
  2. Lending markets that take wstHYPE as collateral see falling collateral value. The Prime docs report that HyperLend values wstHYPE at redemption value, so a pure DEX depeg would not liquidate, but a slash would. This is unconfirmed, because no lender source was read .
  3. Looped Prime positions run at up to 77.77% maximum e-mode loan-to-value, leaving little slashing buffer . A 15% liquidation penalty is reported.
  4. Liquidators may need to sell wstHYPE into a DEX (a community risk review describes the route wstHYPE to WHYPE to USDC, adding hops and slippage), and that DEX held about USD 690K in May 2026 against about USD 65M of collateral . Whether liquidators have other exits is not established, and both figures are from an unverified, five-month-old publication (source reliability: unverified publisher). That collateral is about 35% of today's TVL if it is unchanged since May (a mix of dates, so indicative only).
  • Domains: economic (slashing, thin liquidity, leveraged lender positions) and social (no stress history). No code finding in the reports is tied to this mechanism; the docs' dual-token integration warning is general and is not linked to liquidation here.
  • Likelihood: low (needs a slash or bug first, and the redemption-value oracle behaviour is unconfirmed), but the conditional impact is high. Lender source data and current depth would change the band.
  • Impact: severe for borrowers. Lenders carry bad-debt risk, which sits outside this record.

Vector 4: Single-keeper pause on the restaking escrow

  1. The protocol docs (source reliability: unverified publisher) say a single keeper holding MANAGER_ROLE on the restaking escrow can call setPaused, which blocks deposits, requests, cancellations and execution together, with no maximum duration. The economic report calls this an unconfirmed report .
  2. Restakers also carry slashing exposure and the 14-day cooldown during that time.
  3. The Guardian review of this escrow (2026-10-18, host unverified) lists one informational finding, and the economic report notes that the escrow audit does not mention the pause .
  • Domains: governance (one key, no limit, per the docs), code (audit does not discuss the pause), economic (restaker exit).
  • Likelihood: low while the keeper role and unlimited duration remain docs-only claims; a read of the escrow's roles and pause logic would decide the band. Impact: moderate (limited to restaked funds; the size of the escrow is not measured).

Systemic Risk Factors

  • Single-chain, single-venue dependency (routed to the dependency-risk owner in the economic report): the dependency graph has one edge, from stHYPE to Hyperliquid L1 at share 1.0 . It has no warnings, no circular-backing entries and no downstream-holder entries, and its TVL composition is missing. Every layer a deposit passes through (HyperCore staking, validators, the USDe quote-asset deployer module, and Prime sub-accounts) should be reassessed there. That graph is silent on all of them, and the economic report scores dependency risk as not disclosed.
  • Chain-level: Hyperliquid sets rewards, unbonding limits and slashing. A change in its staking, slashing or CoreWriter behaviour affects every stHYPE account at once . The chain has no L2BEAT stage and no measured exit window.
  • Correlated failure: a slash hits holders, restakers, Prime accounts and wstHYPE collateral in lending markets together, because they all share the same share accounting.
  • Macro: the 7-day target exit and the DEX discount are the stress channel. The Ethena USDe quote asset ties the one active deployer module to a stablecoin issuer. The sources give no data on its failure impact.
  • Points programme: 260M weekly points were fully distributed by 4 June 2026. Whether TVL leaves afterwards is not measured.

Risk Amplifiers

  • No backstop + slow exit + concentrated holders: a loss or a run lands on whoever is left, and a few large holders set the pace. No fund absorbs the first loss.
  • Anonymous team + one multisig + unverified on-chain state: the social report scores identity at its lowest level (team anonymity cap, 4.0), and the governance report cannot confirm signers on-chain. A claim of "4/6" cannot be checked by users.
  • Private audited repository + upgrade on 2026-04-10: the audits are reported as good, but nobody can tie the live bytecode to a reviewed commit. The governance registry documents the proxy upgrade after the Guardian January 2026 review, and the Obsidian April 2026 review covers only part of the code.
  • Pause power + 48-hour upgrade delay + 7-day exit: the same admin can pause burns, and the delay is shorter than a full exit.
  • Own docs as the main evidence: most controls (timelock, 14-day pause cap, roles) are stated only by the protocol docs (source reliability: unverified publisher). A false or stale doc statement would move every governance and economic conclusion at once.

Risk Mitigators

  • Review depth: seven stHYPE-relevant audit reports from four firms over 20 months, with remediation reviews. Several are hosted on unverified hosts (for example Three Sigma, Obsidian and the escrow Guardian report), which lowers confidence in their provenance but not their coverage credit. All critical and high findings for the staking contracts are fixed per the audits . These audits do not cover the older Valantis AMM products, so the AMM audits add nothing here.
  • No rehypothecation in the core token: the docs say stake accounts can only stake to HyperCore, and one third-party review says yield had no leverage; both are unconfirmed outside unverified publishers. No hack is recorded, which does not prove absence.
  • Rebase safety bounds and a pending-slash check on mint and redeem, shown in the Pashov code excerpts.
  • Reduced admin power: threshold raised from 3/5 to 4/6, the address-freeze power removed, and a 48-hour upgrade delay (all docs-stated, not confirmed on-chain).
  • Bounded core pause: a 14-day maximum, documented by the protocol. This is not confirmed in audits or on-chain reads.
  • Redemption-value lender oracle: reported by the Prime docs for HyperLend, not confirmed against the lender.

Contradictions & Inconsistencies

  • Number of audits: the code report counts seven stHYPE audit reports from four firms. The economic report counts eight stHYPE-relevant reports. The difference is due to what each counts, not a data conflict. The Obsidian April 2026 report partly covers Valantis Prime code.
  • Deployer lockup horizon: the docs state 30 to 183-day recalls; Staking Rewards says about 15% of TVL sits in modules with lockups up to 90 days . Both come from unverified publishers; the live figure is unknown.
  • Signer disclosure: the social report records the Staking Rewards claim that "all six signer addresses were publicly disclosed". The governance report says the six addresses are listed but identities and affiliations are not. These fit together: addresses are public, people are not.
  • Timelock ownership of core proxies: the overview says ProxyAdmin ownership sits under a timelock run by the multisig. The governance report says the official registry does not confirm that for the core proxies, only for the escrow; a community review does . Treat the core-proxy timelock as unconfirmed.
  • Bug bounty: the code report says it is not established which statement is current (no bounty for stHYPE versus a USD 200,000 programme on STEX). The social report treats the two as different scopes. The cross-view: no bounty is confirmed for the stHYPE staking contracts themselves.
  • Tone of the audit result versus score: code security reports high review depth, yet the score is capped at 6.0 because audited code cannot be tied to deployed code. This is a measurement gap, not a measured mismatch.
  • Mixed-date numbers: the economic report's DEX depth, collateral and lockup figures are from May–June 2026; the TVL and holder figures are from October 2026. Combined ratios in this report mix these dates.

Overall Risk Profile

stHYPE is a single-chain liquid staking token whose core is reported (by unverified sources) to carry no leverage, with seven audit reports on its staking contracts (several on unverified hosts) and no recorded exploit. Its weak points sit in how the pieces combine. Control is team-held with an anonymous team and unconfirmed on-chain state. Exit depends on the unmeasured instant-queue and on long deployer recalls. Secondary-market depth and holder spread look thin on unverified or dated figures, and loss cover is minimal. Prime traders add a separate seven-day-plus shutdown limit. The most likely failure mode is a stress event in which the queue empties and stHYPE or wstHYPE trades at a discount for days to weeks. The worst case is multisig compromise, or a large deployer-module slash that cascades through wstHYPE lending positions. A comparison with Tier 2 liquid-staking peers could not be assessed, because no peer dataset is in the sources. The component scores (code 6.0, economic 5.5, governance 4.7, social 3.5, track record 6.1) are consistent with the picture above.

Data Gaps

  • Some audit reports were unavailable or could not be read in full.

  • Some source material is unavailable because it is not published.

  • Collection of GitHub repositories stopped at the collection limit; some material may be missing.

  • Current deployment addresses could not be confirmed for some audited contracts.

  • Source summaries cover only the first 80000 characters of 1 document(s); later content was not assessed.

  • No on-chain read of any stHYPE contract on HyperEVM: signers, threshold, timelock, role holders, pause state and stake-account balances are docs statements only. The on-chain tool covers Ethereum, Base and Arbitrum for other Valantis contracts.

  • The audited sthype-contracts repository is private, so audit-to-deployment comparison was not performed. Only the proxy and its implementation have a provider-confirmed source match.

  • No current HyperEVM DEX depth, no instant-queue size, no holder labels, no current lender collateral and bad-debt data. The vector 1 and vector 3 bands depend on these.

  • The size and slashing rules of the USDe quote-asset deployer module are not established, so the slashing magnitude in Risk 3 cannot be narrowed.

  • No historical TVL or APY series, no withdrawal-time statistics, and the quarterly health reports could not be read.

  • Team identity rests on unverified publications. The size of the restaking escrow is not measured.

31 citations · click markers to view evidence

Preview3 · Default / Rows