Cross-Chain Bridge Audit: Scope, Timeline and Cost
Quick answer. A bridge audit covers the contracts on every chain, message verification, relayers and offchain services, configuration such as verifiers and rate limits, and deployment. Prices are rarely published; cost follows code size, chains, runtimes and offchain scope.
Video transcript
A bridge is several systems that have to agree. Ronin, Nomad and Kelp DAO each failed at a different layer.
Scope five layers: the contracts on every chain, message verification, offchain verifiers and relayers, configuration, and deployment.
Kelp DAO was exploited for about 290 million dollars, LayerZero says. The loss began outside the contracts, with poisoned RPC nodes feeding its single verifier.
On Polygon in August 2025, USDT0's contracts were correct. Its migration plan would have left more USDT0 circulating than USDT backing it, and Guardian flagged it before it ran.
Guardian's USDT0 checks match bytecode, confirm privileged addresses, check peers and verifiers on every pathway, and simulate queued transactions.
Audit firms rarely publish prices, and Guardian does not. Cost follows code size, runtimes, chains and routes, offchain scope and fix review rounds.
SEAL roughly estimates two to five weeks for one to four thousand lines. Guardian's published main reviews of one to two thousand lines took a median of eight working days.
Vanguard covers a bridge's spread of surfaces. Two independent teams, one powered by Helix, Guardian's AI auditor. A hundred thousand dollar Defender contest, half the package cost back as offchain credits such as a pentest, and six months of update review.
Read the full article, or send Guardian your repositories, chains and routes for a Vanguard quote.
A bridge is several systems that have to agree: contracts on each chain, verifiers that vouch for messages, and services that carry them.
Major bridge losses came from different layers. Ronin's validators were compromised, a Nomad upgrade made every message valid, and Kelp DAO's route trusted a single verifier.
Here is what to put in scope, how long it takes and what drives cost.
What a Bridge Audit Covers
Five layers, each with its own way to fail:
| Layer | What is reviewed | Example |
|---|---|---|
| Contracts on each chain | Lock, mint, burn and release paths; roles; upgrades | USDT0 review: a USDT quirk that trapped funds |
| Message verification | How the destination decides a message is genuine | Wormhole, 2022: signature check bypassed |
| Offchain services | Verifier signers, relayers, RPC sources, APIs, keys | Ronin, 2022: five of nine validator signatures |
| Configuration | Verifiers, confirmations, peers, rate limits, admin roles | Kelp DAO, 2026: one verifier fed forged data |
| Deployment | Scripts, bytecode, ownership and wiring on every chain | USDT0 Polygon migration plan, fixed before it ran |
Our biggest bridge hacks ranking covers the losses. How bridges get hacked covers each failure mode.
Contracts on Every Chain
On every chain, review:
- Token paths: lock and unlock, or burn and mint.
- Message send and receive, including retries and replay protection.
- Roles: who can mint, pause, upgrade and change settings.
- Upgrade paths and the values set at initialization.
- Pause and rate-limit logic, and what happens to messages in flight.
Chain-specific behavior matters. Guardian found that USDT on Ethereum does not return a boolean from approve(), which made USDT0's multihop contract revert and trap funds.
Other runtimes need reviewers who read them. Guardian's public LayerZero reports alone span Solidity, Rust on Solana and Stellar, DAML on Canton and TypeScript.
Message verification
The code that decides whether to trust a message deserves the hardest review.
Models differ. Wormhole needs 13 of its 19 Guardian nodes to sign. On LayerZero, each app sets required DVNs and a threshold of optional ones, per route.
Two 2022 losses began here. At Wormhole, the Solana contract "didn't correctly verify the address being provided," and 120,000 wETH was minted, per Rekt.
At Nomad, an upgrade set 0x00 as a trusted root, so "all messages were read as valid by default." About $190 million drained, Rekt reported.
Verifiers, Relayers and APIs
The offchain side can fail with the contracts intact. LayerZero says Kelp DAO was exploited for about $290 million in April 2026.
The route relied on one DVN. Per LayerZero, the attacker could not compromise the DVN itself, so it poisoned two RPC nodes the DVN read from.
A DDoS on the other RPCs then forced failover to the poisoned ones.
LayerZero and Kelp disagree on how the route came to rely on one DVN. Our Kelp DAO breakdown has both accounts.
An offchain review asks what would make the system sign or relay something untrue:
- Signing and verification logic in the verifier service.
- Where it reads chain data, and what happens when sources disagree.
- API authentication, access control and input validation.
- Key storage, signer independence and recovery after one node falls.
- Behavior when a job stalls, retries or runs twice.
Configuration and Deployment
A correct contract can still be configured or deployed wrong.
Configuration
LayerZero's docs warn that defaults "may include only a single DVN, and may change without notice." They tell teams to set DVN configuration explicitly.
Review these settings on both chains of every route:
- Required verifiers, optional verifiers and their threshold.
- Block confirmations per source chain.
- Peers, libraries and enforced gas.
- Inbound and outbound rate limits per destination.
- Owner and delegate roles, held by multisigs.
On LayerZero, the delegate "independently controls all Endpoint and security configuration, with no signature required from the Owner." The docs advise a multisig for both roles.
Rate limits cap a forged release. LayerZero's stablecoin extensions set them per destination, in both directions.
They also split pause and unpause roles, so "a compromised pauser key cannot also undo a legitimate security pause."
Deployment
Deployment scripts belong in scope. In Guardian's Citrea Stablecoin Bridge review, 907 of the 1,098 nSLOC in scope were deployment and configuration scripts.
One finding sat in their order: peers were enabled before the destination bridge could mint, so early messages would revert (L-02, resolved).
Then verify each deployment. Guardian's USDT0 checks cover:
- Bytecode matched byte-exact against the audited source.
- Every privileged address confirmed against the intended multisig.
- Peers, DVN sets, confirmation counts and enforced gas for every pathway.
- Every queued Safe and OneSig transaction decoded and simulated.
- Fork tests against live mainnet state.
On Polygon in August 2025, the contracts were correct but the migration playbook was not. Its queued transactions would have left more USDT0 circulating than USDT backing it.
Guardian flagged it, and it was fixed before the transactions ran.
Timeline and Cost
Audit firms rarely publish prices, and Guardian does not. It scopes each engagement from your code and systems.
What drives the cost:
| Driver | Effect |
|---|---|
| Code size (nSLOC) | Review days rise with size |
| Runtimes | Sherlock: Solana Rust carries a 25% to 40% premium over EVM |
| Chains and routes | Each one needs its wiring checked before launch |
| Offchain services | Scoped like a code review, by what has to be read |
| Fix reviews | Plan for at least one remediation round |
Timeline
Published reference points:
- SEAL's rough guide: 1 to 2 weeks under 1,000 lines, 2 to 5 weeks up to 4,000.
- Guardian's published main reviews of 1,001 to 2,000 nSLOC: a median 8 working days.
- Citrea's bridge: main review August 29 to September 3, 2025; fix review September 20.
- Gasolina, LayerZero's DVN service, for USDT0: offchain and API review over two weeks.
- USDT0's Polygon deployment and peer verification: August 5 to 7, 2025.
For the method, see our audit cost estimator and how long an audit takes.
How to Prepare
Send these with your quote request:
- Repositories and a frozen commit for every chain's contracts.
- The chains, runtimes and routes, live and planned.
- A diagram of the message flow, from send to credit.
- The verifier, confirmation and rate-limit settings for each route.
- Source for relayers, verifiers and APIs, and how they are deployed.
- Deployment scripts and the multisig transactions you plan to queue.
- Tests, including fork tests, and your launch date.
For the verifier settings themselves, start with our LayerZero DVN configuration guide.
How Guardian Audits Bridges
Guardian's bridge work includes:
- USDT0: every step of its omnichain rollout verified since February 2025.
- Gasolina: LayerZero's DVN service, reviewed for USDT0; fixes merged upstream.
- LayerZero: contracts across runtimes and the offchain validator stack. See Guardian's public LayerZero reports.
Vanguard, Guardian's enterprise audit, covers a bridge's spread of surfaces:
- Two independent teams on one scope, one powered by Helix, Guardian's AI auditor.
- A $100,000 Guardian-funded Defender contest, typically 20 days and non-blocking for deployment.
- Half the package cost back as offchain credits, for example an offchain automation pentest.
- Six months of update review: up to 2,500 onchain or 10,000 offchain SLOC.
Get a Vanguard quote with your repositories, chains and routes.
Frequently Asked Questions
How much does a cross-chain bridge audit cost?
Audit firms rarely publish prices, and Guardian does not. Cost follows code size, chains and runtimes, the offchain services in scope and fix-review rounds.
What should a bridge audit include?
The contracts on every chain, message verification, offchain verifiers, relayers and APIs, each route's configuration, and the deployment. Ronin's and Kelp DAO's losses began outside the contracts.
How long does a bridge audit take?
It depends on size and surfaces. SEAL roughly estimates 2 to 5 weeks for 1,000 to 4,000 lines. Guardian's Citrea bridge main review ran August 29 to September 3, 2025, with a fix review on September 20.
Do bridge deployments need their own review?
Yes. A correct contract can be wired to the wrong peer, owned by the wrong address or given the wrong verifiers. Check each chain's bytecode, roles, peers and verifiers before a route carries value.
Sources
- Rekt: Wormhole, Rekt (February 3, 2022)
- Rekt: Ronin Network, Rekt (March 29, 2022)
- Rekt: Nomad Bridge, Rekt (August 2, 2022)
- LayerZero: KelpDAO Incident Statement (April 19, 2026)
- Wormhole Docs: Guardians
- Security Alliance (SEAL): What to Expect from a Smart Contract Audit
- Sherlock: Smart Contract Audit Pricing, A Market Reference for 2026 (February 18, 2026)
- LayerZero Docs: Security Stack (DVNs)
- LayerZero Docs: Best Practices for Contract Ownership
- LayerZero Docs: Stablecoin OFT Extensions
- Guardian: Citrea Stablecoin Bridge security review (September 29, 2025)
- Guardian: USDT0 Polygon Deployment and Peer Verification (August 7, 2025)
- Guardian: Public LayerZero reports
- Guardian: Kelp DAO Hack: The rsETH Exploit and What It Means for LRTs
- Guardian: USDT0 cross-chain stablecoin audits case study
- Guardian: USDT0 Deployment and Configuration Verification
- Guardian: USDT0 Gasolina DVN Review
- Guardian: Vanguard
- Guardian: Offchain Automation Pentest
- Guardian: Smart Contract Audit Cost Estimator and Calculator
- Guardian: How Long Does a Smart Contract Audit Take?
- Guardian: Blockchain Penetration Testing Cost
Scope your bridge audit with Guardian
Guardian reviews your contracts on every chain, the verifiers and relayers between them, and each deployment before it carries value.