Improving our Hacking Agent

There have been tons of new developments within the AI Hacking space including talks of Villager, a Chinese hacking tool. What this means for the good guys is that we need this capability. Luckily, some major improvements to our hacking agent have allowed for reduced false positives and much greater capability.

We are excited to announce our validation agent, a one-of-a-kind workflow that takes the proof of concepts developed by Sable and validates them in a novel way, reducing false positives and strengthening the proof of concepts.

When Sable discovers Stored XSS, it reports it as a finding. Our validation agent then kicks off a task with a skeptical view at that finding.

If it is genuine, our agent reports it as “High Confidence”. If not, the agent reports it as a “False Positive.”

Stored XSS validated by Sable’s validator agent

Probably the coolest thing we are releasing is Methodology. This allows you to build a framework our hacking agent uses, automating how you pentest in the exact way you want.

When the user is configuring the target, select the methodology that best fits the assessment type being conducted. This works for Active Directory, web, bug bounties and IoT assessments.

To view our website and try out our product for free: https://vulnetic.ai

To email me directly: danielk@vulnetic.ai