Sensitive Text Redactor
Paste prose or read a text file to find email, Korean mobile number and IPv4 candidates. Add exact custom phrases or manually selected spans. Overlapping or touching candidates become one reviewable span. Check the candidates to mask and replace those spans with [REDACTED]. Detection is incomplete by design; read the source and result before sharing.
Key features
- Candidate detection for email, Korean mobile and IPv4
- Exact custom phrases and manually selected spans
- Merge overlapping/touching spans and choose each group
- Hide candidate values until you choose to reveal them
- Redacted TXT plus value-free offset report JSON
How to use
- Paste text or select a .txt file.
- Choose detectors and custom phrases, then find candidates.
- Review kinds and offsets, revealing values only when needed.
- Select spans and apply masking.
- Download TXT and the value-free JSON report, then manually check for missed sensitive details.
Use cases
- Review contact details in meeting notes before sharing
- Mask phone and IP candidates in sample logs
- Redact a project name or personal name you specify
Frequently asked questions
Will it automatically find every personal detail?
No. It only proposes limited email, Korean mobile and IPv4 forms. Add names, addresses, other-country phones or special formats as custom phrases or manual spans, and check the result yourself.
What happens to overlapping candidates?
Overlapping or touching ranges merge into one group. Selecting that group replaces the entire span with [REDACTED].
Does the report contain original values?
The JSON report contains only UTF-16 start/end offsets, lengths, generic kinds and counts. It has no source or output text. Unmasked source remains in TXT, so review before sharing.
How do custom phrases work?
Enter one phrase per line. Matching is exact and case-sensitive. Up to 40 phrases of 80 characters each are allowed.
Are files uploaded?
A selected UTF-8 .txt file is read by your browser. This tool does not upload source, candidates or output through its own server API.
Privacy
Source text and files are processed in browser memory. The redacted TXT may still contain source text you did not select. The JSON report contains only UTF-16 offsets, lengths, generic kinds and counts, not source values or output text.
Comments & questions