API Endpoints
Overview
The Cerbera app itself is a client of a documented API, and that same API is available for your own integrations, so you can pull data or manage configuration from your own systems instead of clicking through the UI. This supports teams that prefer to automate and want traceability over their configuration, rather than making changes by hand in a portal.
What You Can Do
Read
Pull the AI tool inventory, alerts, and rule matches into your own systems.
Manage
Create and update rules and default policies programmatically.
Read
- Inventory of AI providers, applications, and MCP servers detected across the fleet.
- Alerts and rule matches, with the same privacy rules that apply in the app (metrics by default; bodies only where logging is enabled and released through a data request, see Privacy).
Manage
- Rules: create, update, and remove allow, monitor, redact, and block rules.
- Default policies: set the default behavior for traffic not matched by an explicit rule.
Requesting an exception for a blocked tool or MCP server is handled by email today, not through the API. See Exceptions & Remediation.
Authentication
Requests authenticate with your Cerbera account, using the same authentication the app itself uses, and every action respects the role-based permissions configured in the platform (for example cerbera_user versus employee).
For the exact endpoint reference, request and response shapes, and authentication details, ask your Cerbera security engineer. The capabilities above describe what the API covers today; a self-service, publicly published reference and dedicated API keys for third-party integrations are still being built out.
Infrastructure as Code
Driving configuration through the API is the foundation for managing Cerbera AI as code, with the traceability that brings.
Full infrastructure-as-code support (declarative, version-controlled configuration) is on the roadmap. Design-partner input shapes which resources are exposed first, so tell us what you need.