redact-pii

Redact PII

Redact personal and sensitive identifiers from $ARGUMENTS (or pasted text if no arguments). Replace each match with the listed token. Output the redacted text first, then a one-line counts report.

Patterns

Apply each regex (case-insensitive unless noted) and replace with the bracketed token.

Token	Regex
[EMAIL]	\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,}\b
[PHONE]	(?:\+48[\s-]?)?(?:\d{3}[\s-]?\d{3}[\s-]?\d{3})\b plus \+1[\s-]?\(?\d{3}\)?[\s-]?\d{3}[\s-]?\d{4}\b plus generic \+\d{1,3}[\s-]?\d{4,14}\b
[CARD]	\b(?:\d[ -]?){13,19}\b then Luhn-validate before replacing
[PESEL]	\b\d{11}\b (only when not already matched as [CARD] or [PHONE])
[NIP]	\b\d{3}[-\s]?\d{3}[-\s]?\d{2}[-\s]?\d{2}\b or bare \b\d{10}\b
[PACKAGE_ID]	\b[A-Z0-9]{22}PL\b (24 chars total, ends with PL)
[SESSION_ID]	\b[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}\b
[TOKEN]	\b(?:Bearer\s+)?[A-Za-z0-9+/=_\-]{32,}\b (skip if matches another pattern)
[IBAN]	\b[A-Z]{2}\d{2}[A-Z0-9]{4,30}\b

Procedure

1. Apply patterns in this exact order: EMAIL, IBAN, PACKAGE_ID, SESSION_ID, CARD (Luhn-check), PHONE, PESEL, NIP, TOKEN. Order prevents shorter patterns from cannibalizing longer ones.
2. For CARD: strip spaces and dashes, then run Luhn (sum doubled-from-right digits, mod 10 == 0). Skip replacement if Luhn fails.
3. Track counts per category as you replace.
4. Print the redacted text inside a fenced block.
5. Print one summary line: Redacted: N emails, N phones, N cards, N PESELs, N NIPs, N package IDs, N session IDs, N tokens, N IBANs (omit zero-count categories).

Output template

<redacted text>

Redacted: 3 emails, 2 phones, 1 card, 1 PACKAGE_ID