Explore Security Features

Cencori automatically protects your AI requests. Let's see how it works.

Built-in Protection

Every request passes through multiple security layers:

PII Detection

Automatically detects emails, phone numbers, SSNs, credit cards, and more

Jailbreak Protection

Detects prompt injection attempts and manipulation tactics

Content Filtering

Blocks harmful content, hate speech, and policy violations

Output Scanning

Checks AI responses before they reach your users

Try It: PII Detection

Send a message containing PII and see it get blocked:

test-pii.ts

const response = await cencori.ai.chat({
  model: 'gpt-4o',
  messages: [{
    role: 'user',
    content: 'My email is john@example.com and my SSN is 123-45-6789'
  }]
});

// This will throw a SafetyError!
// Error: Content blocked for safety reasons
// Reasons: ['PII detected: email', 'PII detected: SSN']

Handling Security Errors

handle-security.ts

import { Cencori, SafetyError } from 'cencori';

try {
  const response = await cencori.ai.chat({...});
} catch (error) {
  if (error instanceof SafetyError) {
    console.log('Blocked by security:', error.reasons);
    // ['PII detected: email', 'Jailbreak attempt detected']
  }
}

Security Dashboard

In your project, check the Security section to:

• View security incidents
• Adjust detection thresholds
• Create custom data rules
• Export audit logs

🎉 Congratulations! You now have enterprise-grade AI security without writing any security code.