Crypto glossary
Audit contest
An audit contest is a time-limited public review in which many independent security researchers examine a protocol's code and are paid for valid findings.
In a traditional audit, a protocol hires one firm and a small team reviews the code for a fixed fee. An audit contest opens the code to a crowd instead. For one to several weeks, researchers on a platform such as Code4rena, Sherlock or Cantina study the contracts and submit findings. Judges validate and rank them by severity, and a prize pool is split among those who found real issues, with unique findings worth more than ones many people reported.
Contests bring many different eyes and techniques to the same code and have become a common step before mainnet launches and major upgrades, often alongside a private audit. They have also built public track records for independent auditors, and some platforms add coverage that pays out if a missed bug is exploited.
Quality varies with the size of the prize pool and the code's complexity, and popular contests attract many low-value or duplicate reports that judges must sift through. A contest is a snapshot: code changed afterwards is not covered, and no number of findings guarantees the absence of bugs. Projects still need ongoing bug bounties and monitoring after launch.
Example
Before launching a new lending market, a team funds a three-week audit contest; researchers submit hundreds of reports, judges confirm two high-severity bugs and the team fixes them before deployment.