Kinetiq kHYPE
Score Breakdown
| Category | Weight | Score |
|---|---|---|
| Audits & Historical | 20% | 2.00 |
| Centralization & Control | 30% | 3.30 |
| Funds Management | 30% | 1.75 |
| Liquidity Risk | 15% | 2.00 |
| Operational Risk | 5% | 2.00 |
| Final Score | 2.3 / 5.0 | |
Overview
Kinetiq is a liquid staking protocol for HYPE on Hyperliquid L1. Users stake HYPE and receive kHYPE, a yield-bearing liquid staking token whose redemption value appreciates from staking rewards. kHYPE uses an exchange-rate model (not rebasing) — 1 kHYPE currently redeems for 1.0243 HYPE (verified onchain via StakingAccountant.kHYPEToHYPE(1e18)).
Kinetiq routes stake through a StakingPool contract that manages validator delegation, queue-based unstaking, and fee collection. On 2026-04-07 the StakingPool was upgraded to a Diamond/facet architecture (StakingManagerRouter) introducing a quick-withdrawal path (~36h) and an InstantUnstakePool. Additional products include vaultHYPE / xkHYPE and Kinetiq Markets.
Links:
Risk Summary
Key Strengths
- TVL ~$1.27B with deep DeFi integration (~$499M of kHYPE deposited in lending markets).
- 8 audits from reputable firms (Pashov ×2, Zenith ×2, Code4rena, Spearbit ×3), including a Pashov audit specifically covering the instant-unstake design.
- Active Cantina bug bounty with $1M max and 410 submissions.
- Onchain verifiable staking economics with AccessControl role enumeration; backing reconciles to within ~1,300 HYPE between accountant and actual L1+EVM holdings.
- Diversified delegation: spread across 14 validators with 0% to Hyper Foundation; Kinetiq's share of total network stake is ~2.89%.
- Three exit paths with no exit fee: 7-day standard queue, 36-hour quick withdrawal, and
InstantUnstakePool.
Key Risks
- Standard queue-based unstake (7 days) is the deterministic exit; quick/instant paths are capacity-limited.
- Diamond/facet StakingPool architecture — 7 facets routed via
FacetRegistrycarries higher code surface and integration complexity than a monolithic implementation. - OPERATOR role on StakingPool held by a single EOA (nonce 15,580).
- No timelock on multisig — verified exhaustively onchain (no modules, no guard, no timelock contract). Upgrades can be executed immediately.
- 4-of-8 multisig — threshold is proportionally 50%. All 8 signers are pseudonymous and appear team-associated.
- Hyper Foundation controls 48.13% of network stake — exceeds 1/3 BFT blocking minority (network-level risk, not Kinetiq-specific).
- DEX liquidity is ~$12.6M across HyperEVM DEXes; top pools hold $1.7–5.0M each. Stress-exit scenarios depend on the protocol queue / InstantUnstakePool more than DEX depth.
- Contracts not open-sourced on GitHub (but verified onchain on HyperEVMScan).
- Legal entity ambiguity: "Kinetiq Labs" (Terms) vs "Kinetiq Research" (Privacy/GitHub) with no specific governing law jurisdiction named.
Critical Risks
executeEmergencyWithdrawal()andrescueToken()functions give admin significant power over funds.SENTINEL_ROLEandTREASURY_ROLEon StakingPool are currently unassigned — monitor for grants.
Full Report
Contract Addresses
All contracts are deployed on HyperEVM (Hyperliquid L1). Explorer: HyperEVMScan.
Onchain verified contracts (have deployed bytecode):
| Contract | Address | Type |
|---|---|---|
| kHYPE | 0xfd739d4e423301ce9385c1fb8850539d657c296d |
Proxy (ERC-20 LST) |
| kHYPE Implementation | 0xfe3216d46448efd7708435eeb851950742681975 |
Implementation |
| kHYPE ProxyAdmin | 0x9c1e8db004d8158a52e83ffdc63e37eabea8304c |
EIP-1967 Admin |
| StakingPool (Minter/Burner) | 0x393D0B87Ed38fc779FD9611144aE649BA6082109 |
Proxy |
StakingPool Implementation (StakingManagerRouter) |
0x23fda3b9f944dbeaf2dcd5e7b9f8207b102c933f |
Implementation (upgraded 2026-04-07) |
| StakingPool ProxyAdmin | 0x8194aa9eca9225f96a690072b22a9ad0dd064f64 |
EIP-1967 Admin |
| PauserRegistry | 0x752E76ea71960Da08644614E626c9F9Ff5a50547 |
Proxy |
| PauserRegistry Implementation | 0x85776d01dc61a9685c315443c9e2f449472c2e22 |
Implementation (upgraded 2026-04-07) |
| PauserRegistry ProxyAdmin | 0xd26c2c4a8bd4f78c64212318424ed794be120ea6 |
EIP-1967 Admin |
| StakingAccountant (Proxy) | 0x9209648Ec9D448EF57116B73A2f081835643dc7A |
Aggregator (kHYPE↔HYPE conversion) |
| StakingAccountant Implementation | 0x1bbfe538ccebe1df35c2ebe35203d40cb18fdae0 |
Implementation |
| StakingAccountant ProxyAdmin | 0x7957de697ea42238f53f60ca7ad55b7b03cc0bda |
EIP-1967 Admin |
| ValidatorManager | 0x4b797A93DfC3D18Cf98B7322a2b142FA8007508f |
Delegation engine |
| OracleManager | 0x192826e470bd65FDC2CB472eDd834D096233049b |
Validator performance oracle |
| RewardShareTracker | 0xE5FbA07C7b3CfbC29633f3D3Ab38b36007F35983 |
Reward distribution |
| InstantUnstakePool | 0x665b67793594fc5C251a3C95cbEb4B6245Cd2123 |
Instant-unstake liquidity pool |
| FacetRegistry | 0x4841eCC9Ffb18aEA197E4d1dd1633cEA5e834225 |
Diamond facet registry (7 facets) |
| Governance Multisig | 0x18A82c968b992D28D4D812920eB7b4305306f8F1 |
Gnosis Safe (4-of-8) |
| Treasury Multisig | 0x64bD77698Ab7C3Fd0a1F54497b228ED7a02098E3 |
Gnosis Safe (4-of-7) |
All four ProxyAdmin contracts are owned by the Governance Multisig (verified onchain). The 4-of-8 multisig can upgrade all contract implementations without timelock.
Architecture (2026-04-07 upgrade). StakingPool and PauserRegistry were upgraded in a single transaction (0xe0d05afb…0478a0). StakingPool runs on a Diamond/facet architecture (StakingManagerRouter, 7 facets in FacetRegistry) and exposes a quickWithdrawalDelay() of 36 hours plus an InstantUnstakePool alongside the standard 7-day withdrawal queue.
How Hyperliquid Staking Works (Context)
Hyperliquid staking is validator-based. HYPE is delegated to validators, and rewards depend on validator performance and network-level staking parameters.
Important slashing context (as of February 12, 2026, per official Hyperliquid docs):
- No automatic slashing is implemented ("There is currently no automatic slashing implemented").
- Slashing is "reserved for provably malicious behavior such as double-signing blocks at the same round."
- Validator penalties are jailing-only: jailed validators produce no rewards for delegators, but no principal loss occurs.
- Validators may be jailed by peer quorum vote for inadequate latency/frequency of consensus messages.
- Jailed validators can unjail themselves subject to onchain rate limits.
- Self-delegation requirement: 10,000 HYPE (locked 1 year). Delegation lockup: 1 day.
- Unstaking queue (staking → spot): 7 days. Max 5 pending withdrawals per address.
- Governance can still change staking/penalty rules in future.
Implication for kHYPE:
- Current tail risk is more validator-performance/liveness and queue-liquidity related than immediate automatic slash haircut.
- Governance-introduced slashing would materially change kHYPE risk profile and should trigger reassessment.
Audits and Due Diligence Disclosures
Kinetiq hosts audit reports in a Google Drive folder (linked from https://audits.kinetiq.xyz/).
kHYPE / staking-core relevant audits (verified from Google Drive)
| Date | Auditor | Scope | Link |
|---|---|---|---|
| Mar 2025 | Pashov Audit Group | kHYPE LST | |
| Mar 2025 | Zenith | kHYPE LST | |
| Apr 2025 | Code4rena | kHYPE LST ($35K, 69+ wardens) | Report |
| Jun 2025 | Spearbit | kHYPE LST | |
| Nov 2025 | Pashov Audit Group | kHYPE instant unstake |
Broader protocol audits (additional context)
| Date | Auditor | Scope | Link |
|---|---|---|---|
| Nov 2025 | Spearbit | kmHYPE | |
| Nov 2025 | Zenith | kmHYPE | |
| Jan 2026 | Spearbit | skNTQ |
Architecture complexity: high-moderate. kHYPE relies on multiple upgradeable proxy contracts (kHYPE token, StakingPool, PauserRegistry), which increases integration/control-plane risk compared to single-contract wrappers.
Bug Bounty
- Platform: Cantina
- Max Reward: up to $1,000,000 (Critical severity)
- Scope: kHYPE, StakingManager, StakingAccountant, ValidatorManager, PauserRegistry, OracleManager, OracleAdapter
- Status: Live since September 15, 2025; 410 findings submitted (as of September 2026)
- Link: https://cantina.xyz/bounties/a98129d7-dd15-4c16-b2cb-d8cc42f87de4
Historical Track Record
- Listed on DeFiLlama since July 17, 2025 (~14 months at assessment date).
- Current TVL: ~$1.27B (September 6, 2026, per DeFiLlama) — up ~29% over the last quarter.
- Peak TVL: ~$2.65B (October 4, 2025).
- HYPE price: ~$86 (September 7, 2026, per CoinGecko), ATH $89.60 on September 6, 2026.
- totalSupply (onchain): 12,762,341 kHYPE.
- No Kinetiq entry found in DeFiLlama Hacks database or Rekt News.
- Evolving module set (kHYPE + xkHYPE/skHYPE/kmHYPE + April 2026 StakingPool facet refactor) maintains a moderate-to-high change-risk profile.
Funds Management
kHYPE manages deposited HYPE through a StakingPool contract and validator delegation, not a passive 1:1 wrapper.
Accessibility
- Minting via stake flow is permissionless through app/contract path (
whitelistEnabled()= false onchain). - Three exit paths:
- Standard queue:
withdrawalDelay()= 604,800 sec = 7 days. - Quick withdrawal:
quickWithdrawalDelay()= 129,600 sec = 36 hours. - Instant unstake via
InstantUnstakePool(currently 31,971 HYPE liquid).
- Standard queue:
- Onchain verified parameters:
unstakeFeeRate(): 0 — no exit fee.minStakeAmount(): 5 HYPEwithdrawalPaused(): falsestakingPaused(): false
Collateralization
Onchain state (verified September 7, 2026 at block 45,279,905):
- kHYPE totalSupply: 12,762,341 kHYPE
- StakingAccountant
totalStaked: 52,167,739 HYPE (cumulative deposits-equivalent) - StakingAccountant
totalClaimed: 39,642,151 HYPE (cumulative withdrawn) - StakingAccountant
totalRewards: 547,445 HYPE - StakingAccountant
totalSlashing: 0 - Exchange rate (
kHYPEToHYPE(1e18)): 1.024344 HYPE per kHYPE - Net HYPE backing (
totalStaked − totalClaimed + totalRewards): 13,073,033 HYPE- L1 delegated to validators: 12,718,254 HYPE
- L1 pending withdrawal: 206,985 HYPE (4 active queued positions)
- EVM liquid (StakingPool contract): 117,132 HYPE
- InstantUnstakePool liquidity: 31,971 HYPE
- Sum: 13,074,341 HYPE (matches accountant within ~1,300 HYPE rounding) ✓
- totalQueuedWithdrawals (EVM): 604,543 HYPE
- StakingPool kHYPE held (mid-flight burns): 590,457 kHYPE
Economic backing is fully onchain: ~97.3% delegated to L1 validators, ~1.6% in pending unstaking, the remainder (~1.1%) as EVM-side liquid HYPE (StakingPool balance and InstantUnstakePool) for the instant/quick withdrawal paths. No offchain custodial reserve model is disclosed.
Provability
- Core staking and token accounting are onchain.
- All core contracts (kHYPE, StakingPool, PauserRegistry, StakingAccountant, ValidatorManager, OracleManager, RewardShareTracker, InstantUnstakePool and their implementations) are source-code verified on HyperEVMScan (exact match).
- Key onchain readable functions verified:
totalSupply(),StakingAccountant.totalStaked/totalClaimed/totalRewards/totalSlashing,kHYPEToHYPE,HYPEToKHYPE,withdrawalDelay,quickWithdrawalDelay,unstakeFeeRate, role enumeration viaAccessControlEnumerable. - Exchange rate is derived from onchain accountant state, not admin-set.
- Contracts use OpenZeppelin AccessControlEnumerable. kHYPE supports EIP-2612 Permit.
- kHYPE is not ERC4626 (
convertToAssets,totalAssets,assetall revert).
Liquidity Risk
kHYPE exit routes:
- Protocol unstake queue (primary deterministic exit)
- Onchain
withdrawalDelay()= 7 days;quickWithdrawalDelay()= 36 hours unstakeFeeRate()= 0 — fee-free exits- Queue delay can expand under stress; instant path is capacity-limited by
InstantUnstakePoolliquidity (~32K HYPE currently)
- Secondary market liquidity (per GeckoTerminal + DeFiLlama, September 7, 2026)
Total kHYPE DEX liquidity: ~$12.6M across 80 pools on HyperEVM DEXes.
Top DEX pools:
| DEX | Pair | TVL | 24h Volume |
|---|---|---|---|
| Nest | kHYPE/WHYPE (0.01%) | $5,000,326 | $51,046 |
| Project X | kHYPE/WHYPE (0.01%) | $2,574,749 | $201,297 |
| Ramses v3 (HL) | kHYPE/WHYPE (0.008%) | $1,739,542 | $1,650,980 |
| Project X | vkHYPE/kHYPE (0.2%) | $655,386 | $1,883 |
| Project X | kmHYPE/kHYPE (0.3%) | $293,681 | $11,384 |
| Nest | KNTQ/kHYPE (0.25%) | $290,986 | $146,779 |
Lending protocol deposits dominate external kHYPE usage: ~$322M HyperLend, ~$168M Morpho Blue across 6 markets, ~$4.0M Harmonix, ~$2.9M Felix CDP, ~$2.5M HypurrFi, ~$0.7M Euler v2. Pendle now lists kHYPE markets (~$11M across two pools). These are not exit liquidity.
All trading is DEX-based on HyperEVM. No centralized exchange listings found.
Stress-exit scenarios rely primarily on the protocol's three-tier queue (instant pool → 36h quick path → 7d standard) rather than DEX depth.
Centralization & Control Risks
Governance
Onchain verified governance data (September 7, 2026):
- Multisig address:
Governance Multisig(Gnosis Safe on HyperEVM) - Threshold: 4-of-8 (verified via
getThreshold()).ChangedThresholdevent count over all history: 0. - Version: 1.3.0
- Nonce: 62 transactions executed.
- Timelock: None. Verified onchain — Safe has no modules (
getModulesPaginatedreturns empty array), no guard (storage slot0x4a204f…c34c8is zero), all four ProxyAdmins are standard OpenZeppelin (881 bytes, owned directly by multisig), and noEnabledModuleevents have ever been emitted. - Signer identities: All 8 signers are pseudonymous.
Role structure (verified via AccessControlEnumerable on September 7, 2026):
| Contract | Role | Holder(s) |
|---|---|---|
| kHYPE | DEFAULT_ADMIN | Governance Multisig (4/8) |
| kHYPE | MINTER | StakingPool |
| kHYPE | BURNER | StakingPool |
| StakingPool | DEFAULT_ADMIN | Governance Multisig (4/8) |
| StakingPool | MANAGER | Governance Multisig (4/8) |
| StakingPool | OPERATOR | OPERATOR EOA |
| StakingPool | SENTINEL | (no members — role defined but unassigned) |
| StakingPool | TREASURY | (no members — role defined but unassigned) |
| PauserRegistry | DEFAULT_ADMIN | Governance Multisig (4/8) |
| PauserRegistry | PAUSER | Governance Multisig (4/8) |
| StakingAccountant | MANAGER | Governance Multisig (4/8) — sole authorized manager mapping (StakingPool → kHYPE) |
SENTINEL_ROLE and TREASURY_ROLE exist on the StakingPool but currently have zero members onchain.
Key concern: The OPERATOR role on the StakingPool is held by a single EOA (OPERATOR EOA), not the multisig. This address is a Kinetiq automated bot (nonce 15,580) calling generatePerformance() and updateValidatorMetrics() on a regular basis.
Programmability
- Hybrid onchain system with multiple upgradeable proxy contracts.
- Exchange rate is derived from onchain state (
StakingAccountant.kHYPEToHYPE), not admin-set. - StakingPool has significant admin functions:
pauseWithdrawal(),pauseStaking(),setWithdrawalDelay(),setUnstakeFeeRate(),executeEmergencyWithdrawal(),rescueToken(). Emergency withdrawal and token rescue are powerful admin functions. - Diamond/facet architecture for the StakingPool: a
FacetRegistry(0x4841…4225) routes calls across 7 facet contracts. The implementation (StakingManagerRouter) is source-verified on HyperEVMScan and was preceded by the Nov 2025 Pashov instant-unstake audit.
External Dependencies
Critical dependencies:
- Hyperliquid L1 consensus/liveness.
- Hyperliquid validator performance and staking/slashing rules.
- HyperEVM execution environment.
- DEX liquidity conditions for kHYPE/HYPE exits.
Dependency concentration on Hyperliquid ecosystem is structurally high. HyperEVM is NOT a separate chain — it shares the same HyperBFT consensus as HyperCore. There is no bridge risk between HyperCore and HyperEVM; the risk is pure L1 liveness.
Important: Hyperliquid is a highly centralized chain — Hyper Foundation controls 48.13% of validator stake via 5 validators, well in excess of the 1/3 BFT blocking minority. HYPE staking cannot be considered as safe as ETH staking, where validator set decentralization is significantly stronger (~1M validators, no single entity near blocking minority). This network-level centralization risk is inherited by kHYPE and should be weighed accordingly, even though Kinetiq's own delegations no longer flow to HF validators.
Hyperliquid Validator Set Dependency (Quantified)
Source: Hyperliquid L1 API (POST https://api.hyperliquid.xyz/info, queries validatorSummaries, delegations, delegatorSummary)
Verify validator data: Hyperliquid Staking Portal | Validator Performance | HypurrScan Staking
Network overview (September 7, 2026):
| Metric | Value |
|---|---|
| Total validators | 34 (27 active, 5 jailed, 2 inactive) |
| Total network stake | 440.5M HYPE |
| Active stake | 440.4M HYPE |
| Jailed stake | ~20K HYPE (<0.01%) |
Concentration risk:
- Hyper Foundation operates 5 validators controlling 48.13% of active stake (~212.0M HYPE), still exceeding the 1/3 blocking minority for BFT consensus.
- Top 5 validators (4 HF + Anchorage By Figment) = 51.6% of active stake.
- Top 10 validators = 75.5% of active stake.
- Kinetiq represents ~2.89% of total network stake.
Kinetiq's delegation strategy:
- L1
delegatorSummaryfor StakingPool: 12,718,254 delegated; 206,985 pending withdrawal across 4 positions; ~0 undelegated. - Delegations are spread across 14 validators. Most non-anchor validators receive a uniform ~441,807 HYPE allocation.
- Zero delegations to Hyper Foundation validators.
| Validator | Delegation (HYPE) | % of Kinetiq | Lock Status |
|---|---|---|---|
| HyperStake | 4,158,040 | 32.69% | Unlocked |
| Kinetiq x Hyperion (own) | 3,179,560 | 25.00% | Unlocked |
| Nansen x HypurrCollective | 499,981 | 3.93% | Unlocked |
| Purrposeful x HyBridge x PiP | 461,752 | 3.63% | Unlocked |
| infinitefield.xyz | 442,570 | 3.48% | Unlocked |
| Bitwise | 441,874 | 3.47% | Unlocked |
| Liquid Spirit x Hydromancer x Rekt Gang | 441,832 | 3.47% | Unlocked |
| CMI | 441,807 | 3.47% | Unlocked |
| HypurrCorea - Spacebar x DeSpread | 441,807 | 3.47% | Unlocked |
| USDT0 x Luganodes | 441,807 | 3.47% | Unlocked |
| Hyperdash | 441,807 | 3.47% | Unlocked |
| ASXN | 441,806 | 3.47% | Unlocked |
| B-Harvest | 441,806 | 3.47% | Unlocked |
| Anchorage By Figment | 441,805 | 3.47% | Unlocked |
- 0% of Kinetiq delegations go to Hyper Foundation validators.
- All delegated validators are active with no current jailing exposure.
Slashing/jailing context:
- No automatic slashing is implemented on Hyperliquid (per official docs). However, slashing may be enforced in the future and must be considered a forward-looking risk if Kinetiq delegates to validators that suffer a slashing event (per Kinetiq risk disclosures).
- Jailing = reward cessation only, no principal loss. Jailed validators visible on Hyperliquid Staking Portal.
- Validators can be jailed by peer vote for latency/responsiveness issues (see validator prison docs).
- Unstaking queue from L1 validators: 7 days.
L1 incident history: No Hyperliquid L1 consensus or liveness incidents found in the DeFiLlama hacks database. Four HyperEVM-ecosystem application-level events were recorded (HyperVault $3.6M rugpull, Hyperdrive $773K router exploit, Raga Finance $18.5K exploit, and a "Hyperliquid Malaysia" private-key compromise in August 2026) — none affecting L1 itself.
Operational Risk
- Audit depth is strong for protocol age (8 audits from 4 firms, including coverage of the April 2026 instant-unstake architecture).
- Bug bounty at $1M max is strong and has 410 submissions.
- Team/legal entity: Two entity names are used inconsistently — "Kinetiq Labs" (Terms of Use) vs "Kinetiq Research" (Privacy Policy, GitHub org, footer copyright). GitHub org lists Singapore as location; Privacy Policy references Panama for data transfers. Terms of Use do not name a governing law jurisdiction. No registered address or company registration number is publicly disclosed.
- Known team members (via GitHub commit history on
github.com/kinetiq-research):- Justin Greenberg (@justingreenberg, Twitter: @greenbergz) — primary developer on
f1rewallrepo, PGP-signed commits. - GregTheDev (@0xgregthedev) — Rust/Solidity developer, primary contributor to
hl-rs(Hyperliquid Rust SDK).gregthedev.ethfunded governance Safe Signer 2. - mektigboy (@mektigboy, Twitter: @mektigboy) — self-identified "driver @kinetiq-research" in GitHub bio. Prior experience with sherlock-audit and security audit orgs.
- Justin Greenberg (@justingreenberg, Twitter: @greenbergz) — primary developer on
- Contact: security@kinetiq.xyz (with PGP key at
kinetiq.xyz/.well-known/pubkey.asc), contact@kinetiq.xyz, info@kinetiq.xyz. - No public "About" or "Team" page exists on kinetiq.xyz or docs.kinetiq.xyz. Twitter: @Kinetiq_xyz.
- Contracts are not open-source on GitHub — the Kinetiq GitHub org (
github.com/kinetiq-research) has only SDK/utility repos, no smart contract code. However, all core contracts are source-code verified on HyperEVMScan (exact match). - No public formal verification disclosure found.
Monitoring
Key contracts to monitor:
- kHYPE Proxy:
kHYPE - StakingPool Proxy:
StakingPool - PauserRegistry Proxy:
PauserRegistry - StakingAccountant Proxy:
StakingAccountant - InstantUnstakePool:
InstantUnstakePool - FacetRegistry:
FacetRegistry - Governance Multisig:
Governance Multisig(Gnosis Safe 4-of-8)
1. Governance Monitoring (MANDATORY)
Monitor all privileged role actions and parameter changes for:
- RoleGranted / RoleRevoked events on kHYPE and StakingPool (AccessControl)
- Upgraded events on proxy contracts (implementation changes)
- AddedOwner / RemovedOwner / ChangedThreshold on Governance Safe
Immediate alerts:
- ownership/multisig signer changes
- threshold changes
- implementation upgrades
- emergency pause activations
- parameter changes (withdrawalDelay, unstakeFeeRate, stakingLimit)
2. Backing & Supply Monitoring (MANDATORY)
Track:
kHYPE.totalSupply()(currently 12.76M)StakingAccountant.totalStaked()/totalClaimed()/totalRewards()/totalSlashing()(currently 52.17M / 39.64M / 547K / 0 HYPE)StakingAccountant.kHYPEToHYPE(1e18)exchange rate trend (currently 1.0243)- StakingPool native HYPE balance (liquid buffer, currently 117K HYPE)
- InstantUnstakePool native HYPE balance (instant exit liquidity, currently 32K HYPE)
Alert thresholds:
- backing ratio drift >1% in 24h (unless expected market event)
totalSlashing> 0 (immediate alert — would be the first slashing ever recorded on this protocol)- liquid buffer drops below 50K HYPE
3. Queue Health Monitoring (MANDATORY)
Track:
StakingPool.totalQueuedWithdrawals()(currently 605K HYPE)- L1
delegatorSummary.totalPendingWithdrawal(currently 207K HYPE across 4 positions) - standard (7d), quick (36h), and instant exit utilization
- daily enqueue/dequeue flow
Alert thresholds:
- average unstake latency >10 days sustained
- queue size growth >30% day-over-day
- InstantUnstakePool depletion below 10K HYPE for >24h
4. Market Liquidity Monitoring
Track:
- kHYPE/HYPE and kHYPE/stable liquidity depth on Nest, Project X, Ramses, HyperSwap
- implied NAV discount/premium
- slippage for standard notional buckets
Alert thresholds:
- discount >2% sustained for 24h
- liquidity depth decline >40% day-over-day
5. Hyperliquid Base Risk Monitoring
Track official Hyperliquid updates for:
- validator jailing waves
- staking parameter changes
- any governance introduction of automatic validator slashing
- chain liveness/finality incidents
Reassessment Triggers
- Time-based: Reassess in 3 months
- TVL-based: Reassess if TVL changes by more than 50%
- Any Hyperliquid staking governance change that introduces automatic validator slashing.
- Any kHYPE exploit or emergency pause activation.
- Unstake queue latency >10 days sustained for >72h.
- kHYPE discount >3% sustained for >24h.
- Any privileged role threshold reduction or owner structure downgrade.
- Major contract migration or implementation upgrade (further facet additions/replacements).
- OPERATOR EOA change or role reassignment.
- Assignment of the currently-empty
SENTINEL_ROLEorTREASURY_ROLEon StakingPool. - Resumption of delegations to Hyper Foundation validators (would re-introduce concentration risk).
- Timelock implementation (positive trigger — would warrant score improvement).
Sources
- Kinetiq app: https://kinetiq.xyz/
- Kinetiq docs: https://docs.kinetiq.xyz/
- kHYPE docs: https://kinetiq.xyz/docs/khype
- StakeHub docs: https://kinetiq.xyz/docs/stakehub
- Kinetiq FAQ: https://kinetiq.xyz/docs/faq
- Kinetiq contracts: https://kinetiq.xyz/docs/contracts-and-audits
- Kinetiq audits: https://kinetiq.xyz/docs/contracts-and-audits
- Kinetiq audit PDFs (Google Drive): https://drive.google.com/drive/folders/1T3ZGl6HNmt5LaKwdCmrA9HS7MsXheOys
- Code4rena Kinetiq audit: https://code4rena.com/audits/2025-04-kinetiq
- Kinetiq bug bounty (Cantina): https://cantina.xyz/bounties/a98129d7-dd15-4c16-b2cb-d8cc42f87de4
- CoinGecko kHYPE: https://www.coingecko.com/en/coins/kinetic-staked-hype
- DeFiLlama Kinetiq: https://defillama.com/protocol/kinetiq
- Hyperliquid staking docs: https://hyperliquid.gitbook.io/hyperliquid-docs/hypercore/staking
- Hyperliquid validator prison docs: https://hyperliquid.gitbook.io/hyperliquid-docs/hypercore/staking
- Onchain verification via
castagainst HyperEVM RPC (rpc.hyperliquid.xyz/evm) - Hyperliquid L1 API:
POST https://api.hyperliquid.xyz/info(validator summaries, delegations) - Kinetiq GitHub org: https://github.com/kinetiq-research
- Kinetiq Terms of Use: https://kinetiq.xyz/terms
- Kinetiq Privacy Policy: https://kinetiq.xyz/privacy
- Kinetiq security.txt: https://kinetiq.xyz/.well-known/security.txt
- ENS reverse lookup via
0x3671aE578E63FdF66ad4F3E12CC0c0d71Ac7510C(getNames) - RouteScan API for HyperEVM (chain 999) transaction tracing
- Etherscan V2 API for cross-chain address verification (chain id 999 / HyperEVM)
- GeckoTerminal API for HyperEVM DEX pool data (
api.geckoterminal.com/api/v2/networks/hyperevm) - DeFiLlama yields API for lending market deposits (
yields.llama.fi/pools) - Onchain block reference: 45,279,905 (September 7, 2026)
Score Details
Critical Risk Gates
- No audit -> PASS (5 core audits from 4 reputable firms)
- Unverifiable reserves -> PASS (onchain staking model with readable state)
- Total centralization -> PASS (4-of-7 multisig, not single EOA)
Category Scores
Expand a category to see how it was scored.
Audits & Historical Track Record20%2.00
- 8 audits by reputable firms (Pashov ×2, Zenith ×2, Code4rena, Spearbit ×3) from Mar 2025 – Jan 2026, including coverage of the instant-unstake path.
- Cantina bug bounty at $1M max with 410 submissions.
- ~14 months in production, TVL ~$1.27B (peaked ~$2.65B). No public incidents.
Score: 2.0/5
Centralization & Control Risks30%3.30
Subscores:
- Governance: 4.0 — 4-of-8 multisig (verified onchain), proportionally 50% threshold. No timelock (confirmed exhaustively). Signer independence is questionable (all pseudonymous and team-associated). Powerful admin functions exist (emergency withdrawal, rescue, parameter changes).
- Programmability: 2.5 — Diamond/facet StakingPool architecture (7 facets) carries elevated code surface and integration risk, partially offset by audit coverage. Exchange rate is onchain-derived (via
StakingAccountant). All core contracts source-verified on HyperEVMScan. - External dependencies: 3.5 — Critical single-ecosystem dependency on Hyperliquid L1. Hyper Foundation controls 48.13% of network stake. Kinetiq has zero direct HF-validator exposure, decoupling Kinetiq-specific risk from HF-concentration risk. Kinetiq's network share is ~2.89%.
Centralization score = (4.0 + 2.5 + 3.5) / 3 = 3.33
Score: 3.3/5
Funds Management30%1.75
Subscores:
- Collateralization: 2.0 — 100% onchain collateral (staked HYPE). Backing reconciles cleanly: 13,073,033 HYPE (accountant) vs 13,074,341 HYPE (L1 delegations + L1 pending + EVM liquid + InstantUnstakePool). Exchange rate verified at 1.0243 HYPE/kHYPE. Collateral quality = single-asset (HYPE), high quality within the Hyperliquid ecosystem.
- Provability: 1.5 — Full onchain accountant (
totalStaked,totalClaimed,totalRewards,totalSlashing). AccessControl enumerable. All implementations source-verified onchain (exact match).
Funds management score = (2.0 + 1.5) / 2 = 1.75
Score: 1.75/5
Liquidity Risk15%2.00
- Three deterministic exit paths: 7-day standard, 36-hour quick, and instant pool. No exit fee.
- DEX liquidity is ~$12.6M across 80 pools; largest pool ~$5.0M.
- Most kHYPE in DeFi sits in lending markets (~$304M HyperLend, ~$180M Morpho), which is not exit liquidity.
- All trading liquidity is on HyperEVM DEXes; no CEX listings. The protocol's queue/quick/instant paths carry most of the exit burden.
Score: 2.0/5
Operational Risk5%2.00
- Docs present but client-side rendered (verification difficult).
- Audit cadence and bounty are strong, with coverage of the current architecture.
- Team transparency: 3 identifiable contributors via GitHub history; no public team page.
- No public incident response plan documented.
Score: 2.0/5
Final Score Calculation
| Category | Score | Weight | Weighted |
|---|---|---|---|
| Audits & Historical | 2.0 | 20% | 0.40 |
| Centralization & Control | 3.3 | 30% | 0.99 |
| Funds Management | 1.75 | 30% | 0.525 |
| Liquidity Risk | 2.0 | 15% | 0.30 |
| Operational Risk | 2.0 | 5% | 0.10 |
| Final Score | 2.315 / 5.0 |
Risk Tier
Risk Tier: LOW RISK
Rationale:
- kHYPE is a well-audited LST with significant TVL ($1.27B) and deep DeFi integration.
- Governance is a 4-of-8 multisig with no timelock; signers are pseudonymous and team-associated.
- Diversified delegation: 0% to Hyper Foundation validators, spread across 14 validators; Kinetiq's share of total network stake is ~2.89%.
- Architecture is Diamond/facet (7 facets) — audited and source-verified onchain.
- Three protocol-side exit paths (7d / 36h / instant pool) with no fee; DEX depth is ~$12.6M.
- Hyper Foundation controls 48.1% of network validator stake — structural Hyperliquid risk.
Assessment History
| Date | Score | Notes |
|---|---|---|
| February 17, 2026 | 2.30 | Initial assessment |
| May 19, 2026 | 2.30 | Reassessment: documented the April 2026 StakingPool Diamond/facet upgrade and instant-unstake architecture |
| September 7, 2026 | 2.30 | Reassessment: protocol TVL ~$932M → ~$1.27B as HYPE appreciated ~$49 → ~$86, while kHYPE supply fell 17.4M → 12.76M (net backing 17.72M → 13.07M HYPE) and the exchange rate rose 1.0183 → 1.0243. Hyper Foundation validator stake 54.83% → 48.13%; Kinetiq delegations 16.93M → 12.72M HYPE across 14 validators (network share ~3.96% → ~2.89%). DEX liquidity ~$8.6M → ~$12.6M across 80 pools. Documented the StakingAccountant's fourth ProxyAdmin (0x7957de…b0bda). Bug bounty corrected $5M → $1M max (410 findings). Roles, proxy implementations, timelock, and score unchanged. |