LogScrub Features

LogScrub Features

Free, browser-based PII removal and data anonymization for multiple file formats

LogScrub helps you remove personally identifiable information (PII) and sensitive data from files before sharing them. Everything runs in your browser — no data is ever uploaded to any server.

Supported File Formats

Pattern Matching Rules

LogScrub uses a powerful Rust/WebAssembly engine with 95+ built-in regex patterns to detect sensitive data. All patterns are validated and optimized for accuracy and performance.

Detection Categories

Contact Information

Email addresses, phone numbers (US, UK, International), and usernames.

Network Identifiers

IPv4, IPv6, MAC addresses, hostnames, URLs, and domain names.

Financial Data

Credit cards (with Luhn validation), IBANs, UK sort codes, bank accounts, and cryptocurrency addresses.

Identity Documents

SSN, UK NHS numbers, National Insurance, passports, and international IDs (AU, IN, SG, ES).

API Keys & Secrets

AWS, Stripe, GitHub, OpenAI, Anthropic, Slack tokens, JWTs, and generic secrets.

Dates & Times

ISO dates, US/UK formats, timestamps, and Unix epochs.

Validation

Many patterns include checksum validation to reduce false positives:

  • Credit Cards - Luhn algorithm validation
  • IBANs - MOD-97 checksum
  • UK NHS Numbers - NHS checksum algorithm
  • UK National Insurance - Format and prefix validation
  • Australian TFN - Check digit validation

Custom Rules

Add your own detection rules with full regex support:

  • Custom Regex - Define patterns with standard regex syntax
  • Plain Text - Match exact strings (useful for specific identifiers)
  • Presets - Save and load rule configurations

ML Name Detection

Beyond pattern matching, LogScrub offers optional machine learning detection using a pre-trained Named Entity Recognition (NER) model that runs entirely in your browser.

100% Private
All ML processing happens locally. No data is sent to any server.
Person Names
Detect names that aren't in email or username formats.
Locations
Identify cities, countries, and place names.
Organizations
Find company and organization names.

Technology Stack

  • Library: Transformers.js by Hugging Face
  • Model: BERT-based Named Entity Recognition (NER)
  • Runtime: ONNX format executed via WebAssembly
  • Caching: Model downloaded once, cached in IndexedDB

Available Models

  • DistilBERT NER (~250 MB) - Fast, good accuracy, recommended
  • BERT Base NER (~420 MB) - Best accuracy, slower
  • BERT Base NER (uncased) (~420 MB) - Case-insensitive matching

ML detection complements pattern matching — use both together for comprehensive PII detection.

Key Features

100% Browser-Based
All processing happens locally. Your data never leaves your device.
95+ Detection Patterns
Pre-built rules for IPs, emails, phone numbers, SSNs, credit cards, API keys, and more.
Consistent Replacements
Same value always maps to same placeholder, preserving data relationships.
Custom Regex Rules
Add your own patterns for company-specific identifiers or data formats.
Mapping Export
Download the original→replacement mapping for reverse lookups when needed.
Log Crop
Trim log files to a specific time window. Set a custom range or pick a duration preset to focus on the relevant time period.
High Performance
Rust/WebAssembly engine processes files at native speed.

Common Use Cases

  • Support Ticket Attachments — Remove customer PII before sharing logs with vendors
  • Bug Reports — Sanitize network captures and logs before posting to issue trackers
  • Security Audits — Anonymize data for third-party security assessments
  • GDPR/Privacy Compliance — Redact personal data before data processing
  • Training Data Preparation — Remove PII from datasets used for ML/AI training
  • Documentation Examples — Create sanitized examples from real production data

Ready to anonymize your data?

No installation required. Just open LogScrub and drop your files.

Launch LogScrub