Every part of the platform, in one place — nothing hidden behind a demo call.
Thirteen categories, from the scanning engine down to CLI and SDK access. Every scanning feature below is credit-metered and open on every tier — tiers only gate organizational requirements like SSO, seats, and retention.
See the full breakdown of what's credit-metered vs. tier-gated, and why, on the Pricing page.
Scanning Core
The engine everything else is built around — probe execution against your actual agent, at whatever depth the moment calls for.
Agents & Connections
The unit Orithos scans, and the live endpoints it scans against — described once, reused everywhere.
Findings & Remediation
What a scan actually surfaces, and how you close it — not just a report you read.
Reports & Exports
Evidence that leaves the platform — for a board, an auditor, or your own engineering backlog.
Compliance
Framework coverage that maps every finding to a real control, not just a severity label.
Defenses
Not just finding the gap — closing it, and proving it stays closed.
Adversarial & MCP
The attack surface most competitors don't model at all — the tool layer itself.
Monitoring & Analytics
The always-on view of your posture — never behind a credit meter.
Integrations
Orithos meets you where your code and tickets already live.
Organization, Auth & Security
Everything a real company needs before it lets a whole team in.
Billing & Account
The mechanics behind the pricing page — transparent, and yours to inspect.
Dev Adoption
Orithos in your terminal and your traces, not just your browser.
On the roadmap
Upcoming features
What we're building next. Shipped capabilities stay in the sections above — everything here is actively on our roadmap, not vaporware.
SCIM User Provisioning
Automated user lifecycle management through your identity provider. Provision, deprovision, and sync team members across Okta, Entra ID, and other SCIM directories.
Deployment-Aware Escalation Modeling
Escalation risk analysis driven by your actual agent tool configurations — not generic presets. Model cross-tool attack paths specific to how your agents are wired.
Cross-Provider Security Benchmarks
Aggregate, anonymized security posture benchmarks across LLM providers based on real scan data. Understand how your models compare before you commit.
Ready to probe your AI attack surface?
Join the waitlist for early access and guided onboarding.