🛡️ SafeguardAI - AI Content Moderation & Safety Toolkit

The complete AI content moderation solution for developers. Protect your AI applications from toxic content, detect PII automatically, and ensure compliance with GDPR, COPPA, and HIPAA.

Perfect for: ChatGPT apps, Claude integrations, AI chatbots, user-generated content platforms, and LLM applications.

⚡ Quick Start

npm install safeguard-ai

const SafeguardAI = require('safeguard-ai');

const moderator = new SafeguardAI();

// Basic usage (requires API key unless using mock)
const result = await moderator.checkText("Check this text for safety");

if (!result.safe) {
  console.log('Content flagged:', result.categories);
}

📖 Documentation

• 📘 Getting Started Guide - Complete setup and integration guide
• ⚡ Quick Reference - Cheat sheet for common operations
• 📚 API Documentation - Detailed API reference

🚀 Features

✅ Text Moderation - Detect toxicity, hate speech, violence, sexual content
✅ PII Detection - Find and redact emails, phones, SSN, credit cards
✅ Multi-Provider Support - Currently supports OpenAI, with more coming soon
✅ Custom Rules - Add your own blocked words and patterns
✅ GDPR Compliance - Automatic PII detection for EU compliance
✅ TypeScript - Full TypeScript support included

🌟 The SafeguardAI Advantage

What sets SafeguardAI apart from standard moderation tools:

• All-in-One Safety Stack: Why use three different libraries for PII, Moderation, and Custom Rules? SafeguardAI unifies them into a single, high-performance toolkit.
• Hybrid Processing: Combines cutting-edge AI (for context-aware moderation) with optimized local patterns (for lightning-fast PII detection).
• Privacy-First Design: PII detection and redaction happen locally when possible, ensuring sensitive data never reaches external APIs unless you want it to.
• Zero-Config Start: Get up and running in seconds with sensible defaults, then scale to complex enterprise requirements with custom rules.
• Future-Proof: Built with a provider-agnostic architecture. Switch between OpenAI, Perspective, or Azure with minimal code changes.

💎 How It Helps You

SafeguardAI isn't just a library; it's a protector for your users and your business:

• 🛡️ Protect Your Community: Automatically filter out toxic, hateful, or violent content in real-time, creating a safer space for your users.
• 🔒 Ensure Legal Compliance: Effortlessly meet GDPR, HIPAA, and PCI-DSS requirements by catching sensitive personal data before it's stored or leaked.
• 📉 Reduce Operational Costs: Minimize the need for expensive human moderation teams by automating 99% of regular content checks.
• 🤝 Build User Trust: Show your users that you take their safety and privacy seriously by implementing transparent content safeguards.
• 🚀 Accelerate AI Development: Focus on building your core AI features while we handle the complex logic of content safety and redaction.

Check Text Safety

const result = await moderator.checkText("Your text here");
/* Result:
{
  safe: false,
  flagged: true,
  categories: {
    toxicity: { detected: true, score: 0.89, severity: 'high' }
  },
  piiDetected: [
    { type: 'email', value: '[email protected]', position: [45, 60] }
  ],
  cleanText: "Your text with [REDACTED] instead of PII",
  suggestions: ["Content flagged by OpenAI moderation.", "PII detected in text."]
}
*/

Configure Provider

const moderator = new SafeguardAI({
  apiKey: 'your-openai-api-key',
  providers: ['openai'],
  strictness: 'medium'
});

Add Custom Rules

moderator.rules.addBlockedWords(['badword1', 'badword2']);
moderator.rules.addPattern(/\d{3}-\d{2}-\d{4}/g, 'SSN');

📊 Use Cases

• AI Chatbots - Filter user messages before processing
• Content Platforms - Moderate user-generated content
• Customer Support - Detect PII in support tickets
• Healthcare Apps - HIPAA compliance for medical data
• Financial Apps - PCI-DSS compliance for payment info

🔥 Why SafeguardAI?

| Feature | SafeguardAI | Competitors | |---------|-------------|-------------| | Multi-provider support | ✅ | ❌ | | PII Detection | ✅ | Limited | | Custom rules | ✅ | ❌ | | TypeScript | ✅ | ✅ |

🤝 Contributing

Contributions are welcome! Please read our Contributing Guide for details.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.