doc-redaction-tabular

Redact personally identifiable information from CSV, Excel, Parquet, and DOCX tables.

55|11|Updated Apr 25, 2024
One-click install
npx skills add https://github.com/seanpedrick-case/doc_redaction --skill doc-redaction-tabular
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: doc-redaction-tabular
Source: https://github.com/seanpedrick-case/doc_redaction/tree/main/skills/doc-redaction-tabular
Command: npx skills add https://github.com/seanpedrick-case/doc_redaction --skill doc-redaction-tabular

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill helps you remove personally identifiable information from CSV, Excel, Parquet, and DOCX tables so you can share documents safely without manually scrubbing rows and columns.

Core Features & Use Cases

  • Tabular redaction for multiple formats: Redact table-like inputs including CSV, Excel (.xlsx/.xls), Parquet (.parquet), and Word (.docx) tables.
  • Multiple PII detection modes: Use local detection (spaCy + custom recognizers) or switch to AWS Comprehend when configured.
  • Targeted column and entity control: Choose which columns to process and which entity types to redact, optionally using fuzzy matching and deny/allow lists.
  • Duplicate row handling: Detect and remove duplicate rows via the registered endpoint designed for tabular duplicates.
  • Operational endpoints you can call: Use /tabular_redact when available for short, stateless runs, or /redact_data for the full long signature.

Quick Start

Use the tabular redaction skill to anonymize the columns "Name" and "Email" in your CSV file while replacing detected PERSON and EMAIL_ADDRESS values with the literal text REDACTED.

Frequently Asked Questions about doc-redaction-tabular

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

FAQPage Schema
How do I redact personally identifiable information from CSV and Excel files?

To redact personally identifiable information from CSV and Excel files, you can use column-scoped anonymization to target specific fields and replace detected PII with a designated literal string for safe sharing.

How does tabular data anonymization work with AWS Comprehend?

Tabular data anonymization with AWS Comprehend works by sending file paths through the Gradio API to detect and redact specified PII entity types, offering an alternative to local spaCy recognizers for broader coverage.

Can I choose specific columns and entity types to redact in Parquet files?

Yes, you can specify exactly which columns and entity types to redact in Parquet files, utilizing targeted column scoping and optional fuzzy matching to ensure deterministic redaction outputs.

What is the best way to remove duplicate rows and anonymize DOCX tables?

The best way to remove duplicate rows and anonymize DOCX tables is to use a dedicated endpoint that handles tabular duplicates while applying entity-type selection to redact sensitive table data.

Do I need to configure AWS Comprehend to redact tabular data locally?

No, you do not need to configure AWS Comprehend to redact tabular data locally, as the skill supports local detection using spaCy and custom recognizers for offline data anonymization workflows.

Does Gradio client support sending tabular files for PII redaction?

Yes, the Gradio client supports sending tabular files for PII redaction by transmitting file paths via the handle_file method to execute stateless runs or full long signature redaction processes.