deidentify

Detect and classify PHI columns using locale-aware patterns and column-name heuristics.

243|60|Updated Apr 6, 2026
One-click install
npx skills add https://github.com/Aperivue/medsci-skills --skill deidentify
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deidentify
Source: https://github.com/Aperivue/medsci-skills/tree/main/skills/deidentify
Command: npx skills add https://github.com/Aperivue/medsci-skills --skill deidentify

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires openpyxl, and includes references (resource) components.

What problem does it solve?

De-identify clinical research data before analysis by removing PHI using rule-based, locale-aware patterns and interactive review, without sending data to external models.

Core Features & Use Cases

  • Locale-aware PHI detection: uses country-specific patterns and column-name heuristics to classify fields as PHI.
  • Interactive review and anonymization: researchers approve pseudonymization, date shifting, and suppression in a guided terminal flow.
  • Audit trail and mapping: produces a de-identified dataset plus a secure mapping file and an audit log for IRB compliance.

Quick Start

Run the deidentify tool on your dataset to produce a de-identified copy with an audit trail.

Frequently Asked Questions about deidentify

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

FAQPage Schema
How do I de-identify clinical data without sending PHI to external models?

You can de-identify clinical data locally using rule-based, locale-aware patterns and interactive terminal review to mask PHI without sending data to external models. This approach generates a de-identified dataset and an audit trail for IRB compliance.

What is the best way to generate an audit log for HIPAA compliance when anonymizing research data?

The best way to generate an audit log for HIPAA compliance is to use a guided anonymization flow that documents methodology, stores mappings separately, and produces an audit trail alongside the de-identified data file for IRB review.

Can I interactively review and approve date shifting and pseudonymization for clinical data?

Yes, you can interactively review and approve date shifting and pseudonymization through a guided terminal flow that classifies PHI columns using heuristics and allows researchers to approve suppression actions before generating the final output.

Does locale-aware PHI detection work with country-specific patterns in clinical datasets?

Locale-aware PHI detection works with country-specific patterns by using column-name heuristics and regional formatting rules to accurately classify sensitive fields in clinical datasets before applying masking or pseudonymization.

Why does de-identification require storing the mapping file separately from the dataset?

De-identification requires storing the mapping file separately to ensure security by isolating the pseudonym keys from the masked clinical data, preventing re-identification while maintaining an auditable link for authorized IRB compliance review.