A DeFi lending protocol, scope redacted
A full-stack smart-contract security audit — findings by severity, refusal lanes enforced, operator sign-off on every mitigation.
- Critical finding (mitigated pre-mainnet)
- 1
- Contracts in scope
- 4
- Operator sign-off on mitigations
- 100%
The protocol was preparing a mainnet upgrade across four core contracts and needed an audit that read like engineering, not a marketing PDF. The engagement was a structured security review — static analysis, symbolic paths, and manual review — bound by refusal lanes that forbid automated mainnet execution, fund movement, or performance claims. Every finding is severity-banded; every recommendation waits for the client's security lead to accept, defer, or reject.
- 01Refusal lanes before tooling
Documented the no-list first: no mainnet deploy without signed release, no live fund movement, no yield or return language in the audit surface. The tooling respects the same boundaries the final report does.
- 02Layered analysis surface
Static analysis for known vulnerability classes, symbolic execution on critical paths, manual review on access control and oracle wiring. Findings keyed to contract, line range, and reproducible test case.
- 03Severity-banded reporting
Critical and High findings block the release record until acknowledged or mitigated. Medium and Informational ship with explicit deferral options — nothing silently dropped.
- 04Operator owns every mitigation
Recommendations are drafted; the client's security lead decides. Every decision is logged immutably. The integrity metric is 100% operator sign-off, not finding count.
The audit surfaces risk. The operator decides what ships.
Source notes redacted · Client identification withheld by agreement