🔒 @coffeeandfun/remove-pii

Protect privacy by removing personally identifiable information (PII) from text!

A Node.js library that automatically detects and removes sensitive information from any string — emails, phone numbers, SSNs, credit cards, addresses, and more. Originally built for Helperbird and now used anywhere privacy matters.

Created with ❤️ by Robert James Gabriel at Coffee & Fun LLC.

🎯 Why use this?

For privacy protection

• ✅ Automatically removes sensitive information
• ✅ Prevents accidental data leaks in logs, analytics, and LLM prompts
• ✅ Helps with GDPR, HIPAA, and general privacy compliance
• ✅ Fully customizable per use case

For developers

• ✅ Easy integration — one import, one function call
• ✅ Zero runtime dependencies
• ✅ Pure ESM with TypeScript declarations
• ✅ Detailed analysis and reporting when you need it

📦 Installation

npm install @coffeeandfun/remove-pii

Requires Node.js 16 or newer. The package is pure ESM and ships with TypeScript types out of the box.

🏃‍♂️ Quick start

import { removePII } from '@coffeeandfun/remove-pii';

const text = "John's email is [email protected] and his phone is 123-456-7890.";
const cleaned = removePII(text);

console.log(cleaned);
// "John's email is [email removed] and his phone is [phone removed]."

That's it. Three lines.

🛡️ PII types detected out of the box

💡 URLs are detected but not removed by default — pass { url: { remove: true } } if you want them gone too.

📦 What you get back

Every function takes a text string and an optional options object. The shape of the return value depends on which function you call:

// Simple replacement
removePII(text)                   // → string

// Cleaned text + metadata
removePIIDetailed(text)           // → { cleanedText, removedItems, originalLength, cleanedLength, reductionPercentage }

// Locate without changing
detectPII(text)                   // → { text, detectedItems, hasPII, totalMatches, types }

// Full analysis + risk scoring
analyzePII(text)                  // → { original, cleaned, pii, risk }

// Compliance check
validatePIICompliance(text)       // → { isCompliant, violations, violationCount, riskLevel, riskScore, recommendations }

// Array of strings
processBatch(texts)               // → BatchResult[]

Here's what a removePIIDetailed result actually looks like:

{
  cleanedText: 'Email: [email removed]',
  removedItems: [
    {
      type: 'email',
      count: 1,
      items: ['[email protected]'],
      description: 'Email addresses'
    }
  ],
  originalLength: 23,
  cleanedLength: 20,
  reductionPercentage: 13
}

📚 API reference

removePII(text, options?)

The main function — returns the cleaned string.

const cleaned = removePII("Email: [email protected], Phone: 123-456-7890");
// "Email: [email removed], Phone: [phone removed]"

removePIIDetailed(text, options?)

Like removePII but also tells you what was removed.

const result = removePIIDetailed("Email: [email protected]");
// {
//   cleanedText: "Email: [email removed]",
//   removedItems: [{ type: 'email', count: 1, items: ['[email protected]'] }],
//   originalLength: 23,
//   cleanedLength: 20,
//   reductionPercentage: 13
// }

detectPII(text, options?)

Find PII without changing the text. Each detected item includes positions (start / end / value) so you can highlight matches in a UI.

const analysis = detectPII("Email: [email protected], Phone: 123-456-7890");
// {
//   text: "Email: [email protected], Phone: 123-456-7890",
//   hasPII: true,
//   totalMatches: 2,
//   types: ['email', 'phone'],
//   detectedItems: [...]
// }

analyzePII(text, options?)

The full picture — detection + removal + risk scoring all at once. Great for dashboards.

const analysis = analyzePII("Email: [email protected], SSN: 123-45-6789");
// {
//   original: { text: "...", length: 45, wordCount: 6 },
//   cleaned:  { text: "...", length: 35, wordCount: 6 },
//   pii:      { detected: [...], totalCount: 2, types: ['email', 'ssn'] },
//   risk:     { level: 'medium', score: 13 }
// }

validatePIICompliance(text, options?)

Yes/no compliance check with human-readable recommendations.

const compliance = validatePIICompliance("Email: [email protected]");
// {
//   isCompliant: false,
//   violations: [{ type: 'email', count: 1 }],
//   riskLevel: 'low',
//   recommendations: ['📧 Email detected - Consider using hashed or masked emails']
// }

processBatch(texts, options?)

Runs removePIIDetailed over an array. One bad input won't abort the batch — errors come back as { success: false, error }.

getAvailableTypes()

Lists every built-in PII type with its description and default replacement.

createCustomPattern(type, pattern, replacement, description)

Helper for building a custom-pattern options entry (see Custom patterns below).

🎨 Customization

Basic configuration

Pass an options object to enable/disable types and customize replacements. Anything you don't mention keeps its default.

const options = {
  email: { remove: true, replacement: "[EMAIL HIDDEN]" },
  phone: { remove: false },
  ssn:   { remove: true, replacement: "[SSN REDACTED]" }
};

const cleaned = removePII(text, options);

Privacy levels

🔒 High privacy — lock it all down

const highPrivacy = {
  email:          { remove: true },
  phone:          { remove: true },
  ssn:            { remove: true },
  creditCard:     { remove: true },
  address:        { remove: true },
  passport:       { remove: true },
  driversLicense: { remove: true },
  ipAddress:      { remove: true },
  zipCode:        { remove: true },
  bankAccount:    { remove: true },
  url:            { remove: true },
  dateOfBirth:    { remove: true }
};

🛡️ Moderate privacy — financials only

const moderatePrivacy = {
  ssn:         { remove: true },
  creditCard:  { remove: true },
  bankAccount: { remove: true },
  address:     { remove: true },
  email:       { remove: false },
  phone:       { remove: false }
};

🎨 Custom replacements

const customReplacements = {
  email:   { replacement: "📧 [CONTACT INFO]" },
  phone:   { replacement: "📞 [PHONE NUMBER]" },
  address: { replacement: "🏠 [LOCATION]" }
};

🧩 Custom patterns

Need to redact something the library doesn't know about? Drop in your own regex:

import { removePII, createCustomPattern } from '@coffeeandfun/remove-pii';

const options = createCustomPattern(
  'employeeId',
  /\bEMP-\d{6}\b/g,
  '[employee id removed]',
  'Internal employee IDs'
);

removePII('Ticket raised by EMP-123456', options);
// 'Ticket raised by [employee id removed]'

🔧 Advanced features

📚 Batch processing

import { processBatch } from '@coffeeandfun/remove-pii';

const texts = [
  "Email: [email protected]",
  "Phone: 123-456-7890",
  "Regular text"
];

const results = processBatch(texts);
// [{ index: 0, success: true, cleanedText: ..., removedItems: [...] }, ...]

⚖️ Risk assessment

const compliance = validatePIICompliance(text);

console.log(`Risk Level: ${compliance.riskLevel}`);      // 'low' | 'medium' | 'high' | 'critical'
console.log(`Risk Score: ${compliance.riskScore}`);      // weighted numeric score
console.log(`Recommendations:`, compliance.recommendations);

📖 Inspect available types

import { getAvailableTypes } from '@coffeeandfun/remove-pii';

getAvailableTypes().forEach(t => {
  console.log(`${t.type}: ${t.description}`);
});

🎭 Real-world examples

🧽 Log sanitization

import { removePIIDetailed } from '@coffeeandfun/remove-pii';

export function sanitize(entry) {
  const { cleanedText, removedItems } = removePIIDetailed(entry);
  if (removedItems.length > 0) {
    console.debug(`🔒 Sanitized ${removedItems.length} PII item(s)`);
  }
  return cleanedText;
}

🌐 Express middleware

import express from 'express';
import { removePII } from '@coffeeandfun/remove-pii';

const app = express();
app.use(express.json());

app.post('/logs', (req, res) => {
  console.log(removePII(req.body.message));
  res.sendStatus(204);
});

🔄 API response scrubbing

import { analyzePII } from '@coffeeandfun/remove-pii';

export function scrubResponse(body) {
  const result = analyzePII(JSON.stringify(body));
  return result.pii.totalCount > 0
    ? JSON.parse(result.cleaned.text)
    : body;
}

🧼 Data-cleaning pipeline

import { removePII, validatePIICompliance } from '@coffeeandfun/remove-pii';

function cleanUserData(input) {
  const compliance = validatePIICompliance(input);
  if (!compliance.isCompliant) {
    console.warn(`⚠️ PII detected: ${compliance.violationCount} violations`);
    return removePII(input);
  }
  return input;
}

🧪 Testing

npm test

The suite covers:

• ✅ Every built-in PII type and pattern
• ✅ Edge cases and error handling
• ✅ Performance and consistency
• ✅ Batch processing
• ✅ Custom configurations

🤝 Contributing

We welcome contributions!

1. 🐛 Report issues — found a bug or a missing PII type?
2. 💡 Suggest features — ideas for better privacy protection?
3. 🔧 Submit PRs — code improvements welcome!

Dev setup

git clone https://github.com/coffee-and-fun/remove-pii
cd remove-pii
npm install
npm test

📄 License

MIT — feel free to use this in your projects.

🙏 Credits

Created with ❤️ by:

• Robert James Gabriel — Lead Developer

Originally developed for:

• Helperbird — an accessibility extension making the web more inclusive for everyone

📞 Support

Need help protecting privacy in your applications?

• 🐛 Bug reports: GitHub Issues
• 💡 Feature requests: GitHub Discussions
• 🔐 Security issues: see SECURITY.md

Stay privacy-focused! 🔒