Save my S — AI Prompt Redactor

Redacts sensitive info before it's sent to LLM web UIs. 100% local.

Save my S is a local-only guardrail that catches sensitive info in your prompts before it reaches an AI chat. One absent-minded paste can leak a customer email, an API key, or a card number to an LLM provider forever. Save my S intercepts that moment, shows you a redacted preview, and lets you decide — send the clean version, send the original, or cancel. ━━━ HOW IT WORKS ━━━ • Scans what you type or paste, in real time, inside the browser • Pops a confirmation modal the moment something sensitive is detected • Offers three choices: Send Redacted, Send Original, or Cancel • Shows a live chip near the input so you always know what's been caught • Zero network calls. Zero telemetry. Zero accounts. Your text never leaves your machine. ━━━ WHAT IT CATCHES ━━━ • Email addresses • Credit card numbers (Luhn-validated, no false positives from random digits) • US Social Security Numbers • Phone numbers • IPv4 addresses • Private key blocks (RSA, EC, OpenSSH) • JWT tokens • OpenAI API keys (sk-, sk-proj-) • Anthropic API keys (sk-ant-) • AWS access keys (AKIA / ASIA / AROA / AIDA) • GitHub tokens (ghp, gho, ghs, ghu, ghr) • Google API keys (AIza) • Stripe keys (live & test) • Slack tokens (xoxb, xoxa, xoxp, ...) Plus your own custom regex patterns for internal hostnames, project codenames, or anything else you want to keep quiet. ━━━ WHERE IT WORKS ━━━ On the major AI chat websites — including ChatGPT, Claude, and Gemini. The full list of supported sites is shown in the extension's settings page, where each one can be toggled on or off. A generic adapter also handles sites not explicitly supported. A generic adapter handles the long tail. ━━━ PRIVACY ━━━ This extension does not send any data anywhere. Detection runs entirely on your device using local regex. Settings and counters are stored in your browser. There are no accounts, no sign-ups, no trackers, no analytics. ━━━ WHO IT'S FOR ━━━ • Engineers who paste log snippets and forget the API keys in them • Support teams working in AI chats with customer data • Privacy-conscious anyone using LLMs for everyday tasks