A review starts from the code that is actually deployed: verified source, reconstructed locally so the scope is exactly what holds value. It combines line-by-line manual review with automated analysis — Slither for static analysis, Mythril for symbolic execution, Echidna and Foundry for property testing, Scribble for specifications — and Hardhat compile and regression tests.
The result is a written report: threat model, findings by severity, centralisation notes and guidance for the teams integrating the contract. A clean result is reported as a clean result.