Pendle V2
DAMASCUSYield Trading · Multi-chain · $3B+ TVL · 15 contracts
Public risk assessment — scores are produced with the same methodology as monitored protocols
Security Profile
73
68
62
65
60
78
44
80
75
77
73
68
62
65
60
78
44
80
75
77
Audit History
Bug Bounty Program
Assessment
Novel yield tokenization with complex economic model (D2=68) and heavy oracle dependency (D3=62, OR-001 known). High compositional risk from 30+ SY adapters (D4=65). Governance centralization (D5=60) is a drag. Good maturity and adversarial resilience (zero exploitable findings) lift the score within TEMPERED range.
Dimension Breakdown
How scores work →- 7 access control checks across 8 graphs -- moderate coverage for protocol complexity
- Permissionless market creation increases attack surface
- SY/PT/YT token model with complex mint/redeem flows through 1981 functions
- Reentrancy guards present on core paths
- 2-step ownership transfer (claimOwnership) reduces admin takeover risk
- Novel yield tokenization: PT/YT splitting is unique economic model
- AMM curve (Logit-based) less stress-tested than Uniswap-style
- Implied rate manipulation via AMM state is theoretical attack vector
- Maturity-based expiry creates time-dependent risk profiles
- 19 state writes concentrated in YieldContractFactory -- limited mutation surface
- 151 oracle references in PendlePtLpOracle graph
- Custom TWAP oracle for PT implied rates with known finding OR-001
- PT pricing depends on AMM state (circular dependency risk)
- Oracle manipulation cost varies by market liquidity
- 1 price_feed edge type detected in graph topology
- V2 live since late 2022 (~3.5 years)
- Survived 2023-2024-2025 market cycles including LST/LRT volatility
- Audited by Watchpug, Dedaub, Ackee
- No major exploits on core contracts
- Z-factor: 0.827
- VotingEscrow + GaugeController governance stack (108 + 343 functions)
- CB-004 known finding on VotingEscrow
- Team multisig with no visible timelock on emergency functions
- Centralized parameter control for market creation and fee rates
- setExpiryDivisor, setInterestFeeRate, setRewardFeeRate, setTreasury -- admin-controlled
- Score derived from continuous adversarial security research
- Active development team across multiple chains
- Monitoring infrastructure present
- Bug bounty active on Immunefi
- Incident response untested at scale
- 17 external calls across 8 contracts
- Composes with 30+ yield sources (Aave, Lido, Renzo, etc.)
- Each SY adapter is a trust boundary with unique risk profile
- 28 compound chains found via chain composition (all IRRATIONAL)
- Cross-chain deployments via PendleMsgReceiveEndpoint add bridge-layer risk
- Appears in 4 cross-protocol cascade chain(s)
- Member of 4 dependency cluster(s)
- Score: 77/100 (higher = more isolated from systemic risk)
- Source: cross_protocol_composition.json dependency analysis
- Standard OZ libraries for base contracts
- Each SY adapter adds unique dependency risk
- Proxy patterns used for upgradeability (IMPLEMENTATION_SLOT, ADMIN_SLOT, BEACON_SLOT detected)
- 12 reentry edge types detected -- reentrancy surface exists but guarded
Risk Drivers
Primary risk factors driving this score, ordered by severity.
Adversarial Risk Signals
Observable security posture indicators. These signals reflect publicly verifiable information and responsible disclosure outcomes. No specific vulnerability details are exposed.
Score History & Verification
Score provenance tracking begins with the next reassessment.
On-Chain Data
- Protocol Slug
- "pendle-v2"
- Oracle
- BRORegistry (Base)
- Evidence
- IPFS (pinned)
- Staleness Threshold
- 24 hours
registry.getScore("pendle-v2")Reduce exploitable risk
BlackHart Monitoring provides continuous adversarial analysis, vulnerability detection, remediation support, and verified reassessment when your risk posture improves.