Normal masking (Free)
Detects known secret and PII shapes (API keys, tokens, emails, cards, and more) and blocks them from leaving your browser as raw text.
OmniShield
Mask secrets before they reach AI chat — on your device.
A privacy firewall for ChatGPT, Claude, Gemini, Perplexity, and Cursor. Secrets and PII are masked in your browser before the AI provider ever sees them.
ChatGPTClaudeGeminiPerplexityCursor·On-device Free & Pro
How it works
Three steps. No account required for Free.
- InstallAdd OmniShield from the Chrome Web Store and pin it.
- Chat as usualPaste code, logs, or configs into ChatGPT, Claude, Gemini, or Cursor.
- Secrets get maskedAt send time, matching keys and PII become safe placeholders — on your device.
Pro: context-aware masking
Free finds secrets with patterns. Pro replaces them with format-preserving, context-aware masks so answers stay useful.
Context-aware masking (Pro)
Goes further: replacements preserve format and role (key-like, card-like, identity-like) so the AI can still reason about your request — without ever receiving the real secret.
Built for people who paste into AI
Developers, founders, and teams who use LLMs without leaking keys, customer data, or credentials.
Wire-level masking
Hooks the send path so the raw secret never reaches the model provider's servers — not just a visual blur.
On-device by design
Free and Pro process prompts locally. We do not need your chat content to protect you.
Pro from $24.99
One-time unlock (was $99) — save 75% — for context-aware masking, OCR, healthcare patterns, block mode, and custom rules.
Stop leaking secrets into AI chatbots
Install Free in under a minute. Unlock Pro once for context-aware masking and advanced on-device controls.