Signature verification
Validator set rotation, threshold signing schemes, and key management flaws that let an attacker forge messages or replay them across chains.
// Services / Bridges
Bridges have lost more to exploits than any other category in crypto.
Ronin, Wormhole, Nomad, Multichain. Billions of dollars in losses, across the same category. The reason isn't sloppy code. It's that bridges combine the trust assumptions of two chains with message-validation logic almost nobody else writes. Every bridge invents its own threat model. That threat model gets audited fewer times than the contracts that use it.
Most audit firms treat bridges as smart contract audits: review the L1 contracts, sign off on the relayer interface, ship. The actual attack surface is the system.
// Attack surface
Where the bugs hide.
Message replay
Cross-chain replay protection that breaks under reorganizations or message reordering. The same valid message executed twice drains funds.
Asset accounting
Lock-and-mint logic, supply tracking, and the path between chains where balances can drift if even one assumption is off.
Async finality assumptions
Source chain reorgs and finality gaps that let already-relayed messages get unwound, leaving funds minted on the destination but no longer locked at the source.
Validator and oracle compromise
Trust models that fail if a threshold of validators or oracles collude or get compromised. Most bridge security comes down to this assumption.
Escape hatches and emergency procedures
Mechanisms that exist to recover from failure but introduce new exploit paths if their authorization or sequencing is wrong.
Built by people who've audited cross-chain systems.
Kurt Willis brings deep expertise in layer-2 infrastructure and bridge security from years at Trail of Bits, Spearbit, and Certora. Both founders have audited cross-chain protocols at Offbeat Security. The threat models they internalized — validator compromise, message forgery, async finality bugs, key management failure — are embedded in how the agents prioritize across the full bridge stack.
ZERO COOL
Find what others
miss
For protocol teams and security leaders who want exploit-grade coverage backed by proof, not promises.
Request a demo