Log Snippet Anonymizer & Redactor
Scrub sensitive data from log snippets before you share them. Everything runs in your browser.
What this tool does and does not do — please read
This is a helper that pattern-matches common sensitive data. It is an aid, not a guarantee.
- It does not guarantee it finds every sensitive value. Pattern matching has false negatives (misses) and false positives (over-redaction).
- It cannot know your proprietary or custom secret formats — add them under Custom patterns.
- Pseudonymise is reversible with the on-screen mapping, so it is not legal "anonymisation". Treat the mapping as sensitive.
- Always review the output yourself before sharing a log externally. You remain responsible for what you share.
No warranty on accuracy or completeness. The redacted output and every export contain redacted text only — original values and the pseudonym mapping are never exported.
Up to ~2 MB. Larger input is truncated (you will be told). Nothing is sent anywhere.
Replacement mode
Detect
Custom patterns (your own terms or regex)
No custom patterns. Add a literal term (e.g. a project name) or a regex (e.g. EMP-\d{6}).
Summary
Highlighted spans are what would be redacted, shown over your original text so you can spot misses before copying.
Pseudonym mapping (stays in your browser — reverses the tokens, treat as sensitive)
How to use it
- Paste your log lines, or press Try an example.
- Pick a mode: Pseudonymise (keeps the log analysable) or Mask (opaque).
- Toggle the data types to detect, and add any custom patterns for your own secret formats.
- Press Redact. Check the What was found tab for misses, then Copy or download.
Everything runs locally. The Download files and Copy contain redacted text only — never your original values or the pseudonym mapping.
Frequently asked questions
Do my logs get uploaded anywhere?
No. All detection and redaction runs in your browser with JavaScript. The page makes no network requests with your text (the Content-Security-Policy only allows connections back to this site, and the tool makes none). You can disconnect from the network and it still works.
What is the difference between mask and pseudonymise?
Mask replaces every value of a type with one opaque marker such as [IP] — nothing can be correlated afterwards. Pseudonymise gives each distinct value a stable token (IP_1, EMAIL_2), so the same address reads the same everywhere and the log stays analysable. The pseudonym mapping is reversible, so it is kept in your browser only and never written into the output or an export.
Is this GDPR-compliant anonymisation?
It is an aid, not a guarantee. Pattern matching can miss things and can over-redact. Pseudonymisation is reversible by design, so it is not "anonymisation" in the strict legal sense. Always review the output yourself before sharing.
Why was a 16-digit number not redacted as a card?
Candidate card numbers are checked with the Luhn algorithm before redaction, which rejects most random 16-digit identifiers and cuts false positives. A real card number that is split by unusual separators may be missed — add a custom pattern if needed.