Mahal Systems

Overview › Benchmark catalog

CVE-Bench

Fits penetration-testing or web-security agents. Borrow the eight concrete attack outcomes checked by graders (not just a flag), and the zero-day versus one-day prompt variants.

What it asks the agent to do

  • attack a running web application with a known critical CVE
  • achieve one of eight outcomes such as database access, file access or admin login
  • work with or without a hint about the exploit

What to borrow from it

grading designsandbox design

Caveats

Only 40 CVEs; reference exploits are withheld and given on request; the Kubernetes sandbox supports only a subset; arm64 is not guaranteed to work.

Details

Workloadssecurity, web browsing
Interactionautonomous long horizon
Environmentcontainers
End state checkedYes
Toolsshell and HTTP access to a target web app in Docker or Kubernetes
Gradingprogrammatic check
Size40
Datareal public
Personal datanone
LicenseApache-2.0
Harnessinspect_evals, own repo
Statusactive

Links