sanitizer

Detect secrets, PII, and private context in markdown before release.

2|Updated Feb 23, 2026

Search Tags:#markdown #sanitization #compliance #secrets #redaction #pii #ci

Ready to run

npx skills add https://github.com/MDGrey33/memnyx --skill sanitizer-mdgrey33

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: sanitizer
Source: https://github.com/MDGrey33/memnyx/tree/main/.claude/skills/sanitizer
Command: npx skills add https://github.com/MDGrey33/memnyx --skill sanitizer-mdgrey33

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill protects private repository content from leaking into boilerplate, public releases, or shared documentation by detecting secrets, personal data, private project references, and risky tone before publication.

Core Features & Use Cases

Two-phase workflow that first detects and reports findings, then applies approved fixes.
Scans skills, docs, memory files, and markdown for secrets, PII, private context, and tone risks.
Supports boilerplate, public, and project modes, plus check mode for CI gating.
Use it when preparing a contribution, a shared template, or a release artifact that must stay clean and professional.

Quick Start

Run the sanitizer on the file or folder you want to publish, review the report, and approve any changes only after you are satisfied with the findings.

Frequently Asked Questions about sanitizer

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

FAQPage Schema

How do I sanitize markdown files to prevent secrets and PII leaks before release?▼

You sanitize markdown by running a two-phase detection and approval workflow that scans files for secrets, PII, and private context, reporting findings before applying approved redactions. This prevents risky content from reaching public releases.

How does markdown redaction work in CI gate workflows?▼

Markdown redaction in CI gate workflows uses check-mode reporting to produce deterministic redaction guidance without silent edits. It scans repository files and reports findings, allowing the pipeline to gate contributions based on detected secrets or PII.

Can I use automated PII detection on SKILL.md and CLAUDE.md memory files?▼

Yes, you can use automated PII detection on SKILL.md, CLAUDE.md, and memory files. The scanner detects personal data and private project references before publication, ensuring boilerplate and shared templates remain clean and professional.

What is the best way to redact private project references from shared documentation?▼

The best way to redact private project references is using mode-aware allowlists and denylists during a two-phase scan. This approach identifies risky tone and private context, then applies fixes only after you review and approve the reported findings.

Does this markdown sanitization tool support different publishing contexts?▼

Yes, markdown sanitization supports boilerplate, public, and project modes. These mode-aware settings adjust the detection rules to match your specific publishing context, whether you are preparing a contribution, a shared template, or a release artifact.

Why does the sanitizer report findings instead of silently editing markdown files?▼

The sanitizer reports findings instead of silently editing markdown files to ensure you maintain control over redactions. This two-phase detection and approval process produces deterministic guidance, preventing accidental data loss during secret or PII removal.