Guide
Sensitive Text Redactor
Find and mask common emails, phone numbers, IP addresses, payment-card-like numbers, URLs, and your own specified terms without uploading the text.
Does the text leave this browser?
No. Pattern detection and replacement run in the current tab without a submission request.
Opens the free tool. Your selected files stay in this browser.
Clear answers
Create a shareable text copy with common identifiers masked locally and a clear receipt of what the rules detected.
Find and mask common emails, phone numbers, IP addresses, payment-card-like numbers, URLs, and your own specified terms without uploading the text.
-
01
Does the text leave this browser?
No. Pattern detection and replacement run in the current tab without a submission request.
Pattern-based local detection
-
02
Can it find every name or address?
No. This version uses explicit patterns and custom terms, not an AI model. Always read the complete result before sharing it.
Custom terms and category controls
-
03
Are card-like numbers validated?
The tool uses length, formatting, and a Luhn checksum to reduce false matches, but the result still requires manual review.
Redaction receipt and editable output
Real-life uses
Practical ways to use Sensitive Text Redactor
These examples show where the tool can help and what still needs a deliberate human check before the result is used or shared.
Sharing a support log
Mask emails, phone numbers, IP addresses, URLs, card-like numbers, and known custom terms before pasting diagnostic text into a ticket.
Preparing a message transcript
Remove common identifiers and manually review names, addresses, account references, and contextual clues before wider circulation.
Publishing a sample dataset
Use pattern detection as one review aid, not as proof that free-form text has been anonymized.
Worked example
Preparing an error log for a public forum
The log contains a useful error trace along with an email address, local IP, support URL, and internal project name.
Before
Technical evidence includes identifying values
- The error sequence should remain readable
- Patterned identifiers and project-specific terms should be removed
After
Automatic matches and context are reviewed
- Select detection categories and add the internal project name as a custom term
- Read the entire edited output and manually remove anything the rules missed
What you receive: an editable masked text copy and a receipt of rule-based matches; it is not guaranteed anonymous.
Four local steps
Can it find every name or address?
- 01
Pattern-based local detection
Paste text without including anything you cannot safely display on your current device.
- 02
Custom terms and category controls
Choose detection categories and add optional custom terms.
- 03
Redaction receipt and editable output
Create the redacted copy and review every replacement.
- 04
Pattern-based local detection
Edit remaining details manually before copying or downloading.
Final check
Check the result before you use it
Verification belongs in the workflow. Review the downloaded result itself rather than relying only on the preview inside the tool.
- Read the complete redacted output from beginning to end.
- Search for known names, identifiers, domains, paths, and custom terms.
- Check whether surrounding context can still identify a person, account, customer, or organization.
Clear answers
Are card-like numbers validated?
The tool uses length, formatting, and a Luhn checksum to reduce false matches, but the result still requires manual review.
- Pattern matching can miss sensitive context and can flag harmless text.
- Names, addresses, account identifiers, and unusual formats may require manual review.
- This is not AI classification, legal redaction, or a guarantee that the result is anonymous.