Secure · Detection Rules
Detection Rules
The catalog of built-in risk detectors, plus custom rule authoring with CEL expressions, testing, and AI suggestions.
The Detection Rules tab catalogs the detectors that power guardrail policies. Open it from Secure > Guardrails in the dashboard and select the Detection Rules tab. Old links to the standalone Detection Rules page redirect to the tab.
Access requirements
Section titled “Access requirements”Viewing this page requires the org:admin scope. Access is included in the default Admin role but not the Member role.
Built-in rules
Section titled “Built-in rules”Built-in detectors are grouped by category: secrets, financial data, PII, government IDs, healthcare data, shadow MCP, destructive tool use, account identity, and prompt injection. Expanding a category lists its individual rules; clicking a rule opens a detail sheet, and rule detail views are deep-linkable for sharing.
Custom rules
Section titled “Custom rules”Custom rules extend detection beyond the built-ins:
- Write the match logic as a CEL expression in an editor with syntax highlighting and worked examples
- Assign the rule a category and a severity level
- Rule IDs are prefixed
custom.and validated on save
An AI suggestion action drafts a rule from a plain-language description of what to detect.
Testing rules
Section titled “Testing rules”The test flow runs a rule against a selected chat session and shows exactly what it matches, so a rule can be validated against real traffic before it’s attached to a policy.