Guardrail Policy

Guardrail policies are customizable filters that monitor and analyze your agent traffic in real-time to detect and mitigate agentic security threats. These policies protect your AI agents and MCP endpoints by identifying and blocking malicious activities, such as prompt injections, tool abuse, and attacks outlined in the OWASP Core Rule Set including Cross-Site Scripting (XSS) and SQL Injection. You can configure policies to enforce geofencing, block traffic based on IP CIDR ranges, and control tool access patterns.

To create guardrail policies, head to Create Guardrail Policies.