security

To prevent SQL injection and XSS in agent-native apps, enforce strict input validation using defineAction with Zod schemas. This ensures user data is safely sanitized and scoped across action, route, and component layers before processing.

The best way to handle secret management and prevent credential leakage is using proper credential storage and the framework's security primitives. Guarding fetch patterns and safe data scoping isolate credentials from exposure across app layers.

Zod schema validation secures user data processing by enforcing strict input boundaries within defineAction. This mechanism validates incoming data types and structures, preventing insecure inputs and data breaches before the application logic executes.

Yes, you can apply these secure coding practices to existing routes and components. The guidelines enforce data isolation and defense against SSRF across your app's action, route, and component layers, ensuring safe data access and preventing insecure deployments.

You need guarded fetch patterns to prevent SSRF whenever your application makes external network requests based on user input. Applying these security primitives restricts unauthorized server-side resource access and protects against data leakage.