pii-detector-redactor

Detect and redact PII from documents with configurable styles and audit logs.

Updated Feb 20, 2026
One-click install
npx skills add https://github.com/garethdaine/agent --skill pii-detector-redactor
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pii-detector-redactor
Source: https://github.com/garethdaine/agent/tree/main/skill-library/skills/pii-detector-redactor
Command: npx skills add https://github.com/garethdaine/agent --skill pii-detector-redactor

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill protects sensitive information by automatically detecting and redacting personally identifiable information (PII) from documents and agent outputs, ensuring GDPR compliance and preventing data breaches.

Core Features & Use Cases

  • Comprehensive PII Detection: Identifies a wide range of PII categories including names, addresses, emails, phone numbers, national insurance numbers, passport details, and bank account numbers.
  • Configurable Redaction Styles: Offers multiple redaction methods such as full replacement, partial masking, consistent pseudonymisation, or complete removal.
  • Audit Logging: Generates detailed audit trails for all detection and redaction actions, crucial for GDPR accountability.
  • Use Case: Before sharing a customer support transcript externally, use this Skill to automatically redact all names, email addresses, and phone numbers to protect customer privacy.

Quick Start

Use the pii-detector-redactor skill to redact all personal information from the provided document 'customer_feedback.txt'.

Frequently Asked Questions about pii-detector-redactor

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automatically redact personally identifiable information from documents for GDPR compliance?

Automated PII redaction detects sensitive data like names, addresses, and national insurance numbers, then applies masking or removal techniques to ensure GDPR compliance. It scans structured and unstructured text, classifies categories with confidence scores, and outputs redacted documents with audit logs.

What is the best way to anonymise customer support transcripts before sharing them externally?

Anonymising customer transcripts involves detecting PII such as emails, phone numbers, and names, then applying configurable redaction styles like partial masking or pseudonymisation. This replaces sensitive data with consistent placeholders, making the transcript safe for external sharing while retaining context.

Can I configure different redaction styles for specific PII categories like bank account numbers?

Yes, redaction styles are configurable per PII category. You can choose full replacement, partial masking, consistent pseudonymisation, or complete removal for specific data types like bank account numbers, ensuring flexibility based on your data protection requirements.

Does PII detection provide audit logging for GDPR accountability?

Yes, the process generates structured JSON audit logs detailing all detection and redaction actions. These logs include detection summaries, flagged items for review, and confidence scores, providing the necessary accountability trails required for GDPR compliance.

How does contextual analysis improve PII detection in unstructured documents?

Contextual analysis improves PII detection by examining the surrounding text to accurately identify and classify sensitive information in unstructured documents. It assigns confidence scores to detected items, reducing false positives and ensuring precise redaction of names, addresses, and financial details.

What types of personally identifiable information can be detected and redacted?

PII detection identifies names, addresses, emails, phone numbers, national insurance numbers, passport details, and bank account numbers. It processes both structured and unstructured data, categorising each detection with confidence scores for accurate redaction.