PII Redactor

Redact Personal Info Before You Share It

Paste a message, log line, or support ticket and this masks the personal data inside it — email addresses, phone numbers, Social Security numbers, credit-card numbers (Luhn-validated), and IP addresses (v4 and v6), with optional best-effort name detection. Every match becomes a plain [REDACTED] or a numbered tag like [EMAIL_1], and you get a live count of what was caught in each category.

Because it runs on plain pattern-matching entirely in your browser, the text you scrub never leaves your device — nothing is uploaded, logged, or sent to a server. That is the whole point: you can clean up a customer ticket or a stack trace before pasting it into ChatGPT, a shared doc, or a public bug report without leaking the data you are trying to protect in the act of protecting it.

Detection is heuristic, not magic. It reliably catches structured data like emails, cards, and SSNs; name detection is a best-effort guess that will miss some names and occasionally flag a place or product. The highlighted preview shows exactly what matched, so you can eyeball it before you trust the output — always review, never rely on it blindly.

Common Use Cases

• Prepping prompts for AI: Strip customer emails and phone numbers out of a support thread before pasting it into ChatGPT or Claude for a drafted reply.
• Sanitizing bug reports: Remove real IP addresses and account numbers from logs and stack traces before attaching them to a public issue.
• Sharing sample data: Turn a real record into a safe example with typed tokens — every [EMAIL_1] stays consistent throughout, so the structure survives.
• Forwarding copied text: Clean up text pulled from a ticket, invoice, or form before you send it on to a vendor or a teammate.
• Card and SSN hygiene: Catch Luhn-valid credit-card numbers and formatted SSNs that slip into chat transcripts or email threads.
• First-pass compliance: Give a document a quick scrub of obvious personal identifiers before a wider internal share.

Privacy: This tool runs entirely in your browser using local pattern-matching — your text is never uploaded, stored, or transmitted anywhere, and there are no network calls at all. Because detection is heuristic, treat it as a strong first pass and always review the highlighted matches before relying on the result.