See the masking, live
Type anything. Sensitive values are detected and replaced with tokens before they'd ever reach an LLM — then restored in the reply. This runs entirely in your browser: your text never leaves this page.
View the open-source engine on GitHub →Detects: EmailPhoneCredit cardSSNIP addressAPI keyUUID
What the provider sees
Hey — email me at ⟦EMAIL_1⟧ or call ⟦PHONE_1⟧.
Charge card ⟦CREDIT_CARD_1⟧, SSN ⟦SSN_1⟧.
Prod server ⟦IP_1⟧, trace ⟦UUID_1⟧.
6 sensitive values masked before anything leaves your machine.
- EMAIL maria@example.com → ⟦EMAIL_1⟧
- PHONE 415-555-0132 → ⟦PHONE_1⟧
- CREDIT_CARD 4242 4242 4242 4242 → ⟦CREDIT_CARD_1⟧
- SSN 123-45-6789 → ⟦SSN_1⟧
- IP 10.0.0.42 → ⟦IP_1⟧
- UUID 7c9e6679-7425-40de-944b-e07fc1f90ae7 → ⟦UUID_1⟧
This is just the masking step — running locally.
Connect your own Claude or OpenAI key to get the full round-trip: mask → your LLM → rehydrate, inside the app.