Skip to main content

Welcome to Cybrium

Cybrium is an AI-powered penetration testing and security platform. It combines twelve scan types, an autonomous adversary engine, and a continuous exposure-management workflow into a single workspace — so your security team can shift from "scan → triage → forget" to "discover → exploit → fix → verify".

This documentation is organised for tenant users — the teams running scans inside their Cybrium workspace. Platform administrators managing the control plane have a separate section (Sprint 32).

At a glance​

AreaWhat you get
12 scan typesNetwork · SAST · DAST · SSL/TLS · SBOM · Secrets · Traffic · Cloud · Kubernetes · NHI · M365 · API
Adversary engineAutonomous MITRE-ATT&CK-mapped campaigns, Black/White/Grey/Red/Blue box
Distributed originsAttack from regions across 5 continents — Native, WireGuard, Proxy, Tor exits
Live viewerWebRTC desktop + terminal stream of the attacker VM as it runs
CTEM pipelineExposure scoring, SLA policy, AI Fix Bot, autonomous remediation
ComplianceCIS · HIPAA · PCI-DSS · SOC 2 · OWASP · NIST · ISO 27001 mapping

Where to start​

Brand-new to Cybrium? Follow the Getting Started path.

Already signed in? Jump to your scan type:

Screenshot: Cybrium workspace overview

Conventions used in this doc​

Tenant admin only

Callouts with this icon mark features that require the admin or owner role within your tenant. Regular members see read-only versions.

Requires authorisation

Callouts with this icon denote actions that run against real systems and require a signed authorisation document + liveness check before starting.

Plan-gated

Callouts with this icon mark capabilities scoped to Pro or Enterprise plans. Starter tenants see the feature but cannot launch it.

Need help?​

  • In-product chat is always on — click the purple chat icon in the top right of your workspace.
  • Email: support@cybrium.ai
  • Docs feedback: open an issue at github.com/cybrium-ai/docs or click the Edit this page link at the bottom of any article.