We don't auto generate issues exploits but rather find the already existing exploits and break them further to test the full depth of the vuln. We use some aspects of graph DB but its not quite the same thing. This differs from Mend and CodeQL because they focus on deep semantic analysis or SCA, We use parallel detection systems for hybrid, holistic analysis by combining advanced static testing, execution modeling, and ML on test data to improve bug breadth, path feasibility, and alert prioritization aka. prove deep rooted issues other tools are not trained to find.
And yes it does integrate in line with the code and trace exec flow. Would you wanna try it out and see what it can help you find? It runs locally so nothing leaves your system
reply