We benchmarked our AI audit engine against Jito Restaking — a 9,000-line Rust/Solana protocol that was the subject of four professional audits and a $150,000 Immunefi bug bounty competition. This is RedVolt's first Rust/Solana benchmark, and the results exceeded our expectations.
100% critical and high vulnerability detection. 90.3% OWASP Top 10 coverage. Real numbers, real targets, no cherry-picking. Here are our AI pentest engine benchmark results against OWASP Juice Shop.
We benchmarked our AI audit engine against a real Code4rena contest. Every high and medium finding detected. Every PoC verified by Foundry. 100% severity accuracy. Here are the full results.
22 Solidity files. 2,332 lines of code. 6 high-severity bugs. Our AI engine found every single one — outperforming 90% of the 144 human wardens in the original Code4rena contest.
Reentrancy, flash loan exploits, share inflation, gas DoS, access control — our AI engine caught every vulnerability class across 7 battle-tested CTF challenges. With proof-of-concept for each.