Email and Cyber Security · #74
Online Log Masking Tool
Masks personal and secret data on your device before you share logs with third parties.
- Data is processed in your browser and not sent to a server
Tool
What does this tool do?
Logs sent to vendor support, forums or consultants often contain customer IPs, email addresses, session tokens and even passwords. The tool recognises these values; in pseudonym mode the same value gets the same name everywhere (IP-1, EMAIL-1) so the log stays analysable. Card numbers are validated with Luhn, IBANs with mod-97 and Turkish ID numbers with their checksum before masking, reducing false positives.
How to use it
- Paste the log text.
- Choose data types to mask and between consistent pseudonyms or plain [TYPE] labels. Add extra words (server, customer names) one per line if needed.
- Press Run; copy the masked text and review it before sending.
Fields
- Log text
- Up to 2 million characters. Data does not leave the browser.
- Use consistent pseudonyms (IP-1, EMAIL-1)
- IPv4 and IPv6 addresses
- Email addresses
- Passwords, tokens, API keys, private keys
- Usernames (DOMAIN\user, user=)
- MAC addresses
- Card numbers (Luhn validated)
- IBAN (mod-97 validated)
- Turkish ID number (checksum validated)
- Extra words to mask
- One word per line; case-insensitive.
Worked example
| Input | Value |
|---|---|
| Log text | |
| Use consistent pseudonyms (IP-1, EMAIL-1) | ✓ |
| IPv4 and IPv6 addresses | ✓ |
| Email addresses | ✓ |
| Passwords, tokens, API keys, private keys | ✓ |
| Usernames (DOMAIN\user, user=) | ✓ |
| MAC addresses | ✓ |
| Card numbers (Luhn validated) | ✓ |
| IBAN (mod-97 validated) | ✓ |
| Turkish ID number (checksum validated) | ✓ |
Expected result: 11 values masked: the same attacker IP is IP-1 on both lines, the email EMAIL-1, password and Bearer token SECRET, plus the MAC, the SSH user names admin and ayse, and the CONTOSO\ali.yilmaz username.
Usernames in free text such as 'invalid user admin' and 'password for ayse' do not match a pattern and were not masked; add them under 'Extra words' if required. The server name FS01 also remains visible. Always do a final visual check before sending.
Second scenario
| Input | Value |
|---|---|
| Log text | |
| Use consistent pseudonyms (IP-1, EMAIL-1) | – |
| IPv4 and IPv6 addresses | ✓ |
| Email addresses | ✓ |
| Passwords, tokens, API keys, private keys | ✓ |
| Usernames (DOMAIN\user, user=) | ✓ |
| MAC addresses | ✓ |
| Card numbers (Luhn validated) | ✓ |
| IBAN (mod-97 validated) | ✓ |
| Turkish ID number (checksum validated) | ✓ |
Expected result: In label mode the Turkish ID, card and IBAN become [TCKN], [CARD], [IBAN]; order number 4711, which fails the checksums, is kept.
Method and assumptions
Rules are applied in order: secrets, email, IBAN, card, Turkish ID, MAC, IP, username, extra words. Already masked values are not processed again.
Card: 13–19 digits and Luhn; IBAN: country code + mod-97 = 1; Turkish ID: 11 digits, first digit not 0, 10th and 11th digit algorithm. IPv6 is masked only if it parses as a valid address.
Using the result and troubleshooting
Pseudonym mode keeps relationships for troubleshooting; prefer label mode for logs handed over as a permanent archive. Masking is not an anonymization guarantee; context (time, device name) may still identify a person.
Limits and common mistakes
- Personal names, postal addresses and phone numbers in free text are not recognised. No field-name-based masking inside multi-line JSON/XML.
Frequently asked questions
Is a masked log anonymous under GDPR?
Not always. Pseudonymised data remains personal data if it can be linked to a person via a mapping table or context. Consider the purpose and the recipient.
Related tools and service
- Security Incident Timeline BuilderMerges EDR, firewall, SIEM and team notes into one sorted, time-zone-corrected timeline.
Related service: SIEM Setup and Security Logs