scrubfile

Scrub PII from PDFs, images, and DOCX files. Local-only. One command.

$pip install scrubfile

GitHub PyPI Docs

Why scrubfile?

🔒

100% Local

No cloud APIs. No data leaves your machine. Zero network calls after model download.

📄

Multi-Format

PDF, PNG, JPG, TIFF, BMP, DOCX. One tool handles them all.

🧠

Auto-Detect PII

Names, SSNs, emails, phones, addresses, credit cards, and 20+ entity types via Presidio + spaCy.

⚙️

Permanent Redaction

Text removed from PDF content stream. Not a visual overlay — the data is gone.

🤖

MCP Server

AI agents (Claude, Cursor) can redact documents directly via Model Context Protocol.

📦

JSON + CLI + API

Machine-readable output. Python API. CLI with rich output. Built for automation.

Simple by design

            # Redact specific terms

            $ scrubfile document.pdf -r "John Doe" -r "123-45-6789"

            # Auto-detect all PII

            $ scrubfile document.pdf --auto

            # Python API

            from scrubfile import redact

            result = redact("document.pdf", auto=True)

How it compares

	scrubfile	Adobe Acrobat	Google DLP	Presidio
Local-only	Yes	Yes	No	Yes
Multi-format	PDF, images, DOCX	PDF only	Text/images	Text only
CLI	Yes	No	No	No
Auto-detect PII	Yes	No	Yes	Yes
Agent-ready (MCP)	Yes	No	No	No
Free	Yes	$240/yr	Pay per call	Yes