Start from any origin.
See direct recorded routes from your runner or another origin. Understand the view from each point.
Your AI security workspace. Investigate in conversation, connect the evidence, and see your attack surface take shape.
Copy, paste, and start investigating.
curl -fsSL https://raw.githubusercontent.com/Kerkov-Security/kerkov/main/install.sh | bashkerkov and choose your model.Hosts, containers, services, and recorded access routes. Bring the engagement into one view, then inspect what connects it.
INTERACTIVE PREVIEW · SAMPLE DATAYour engagement, connected.
See direct recorded routes from your runner or another origin. Understand the view from each point.
Inspect services, ports, and linked evidence. Observed routes and unverified declarations stay distinct.
Group host resources, recognize container endpoints, and explore the inventory saved to your engagement.
Work through an engagement in conversation. Your tools, specialist agents, inventory, and findings share the same context.
Map the staging API and review its access controls. Keep the findings linked to evidence.
I’ll inspect the application, record the connections I can verify, and keep active testing behind your approval.
readInspect application routesledgerSave resources and connection evidenceapprovalReview before active testingRead code, run commands, make HTTP requests, and drive a browser. Bring OpenGrep, ZAP, Syft, and Grype into the workflow.
Delegate focused work to agents with scoped tools. Follow their activity and bring structured reports back to the engagement.
Choose your operating mode and approval policy. Save the engagement and pick up where you left off.
Use a hosted provider or a local model through Ollama.
The engagement stays with you.
One command to install. A conversation to get started.
curl -fsSL https://raw.githubusercontent.com/Kerkov-Security/kerkov/main/install.sh | bash