Support

A few things to check:

• Open the toolbar popup. Does it show "Active on [site]"? If it says "Paused," click Resume.
• Open Options → Sites. Make sure the site you're on (e.g., ChatGPT, Claude) is toggled on.
• Open Options → Detection rules. The category you expect to trigger must be enabled. Pro categories are locked on the free tier.
• For X / Grok: the extension only acts when you're at x.com/i/grok. It silently ignores tweet drafts, reply boxes, and DMs.

Two options:

• "Ignore for this session" on the modal row — temporary; the snippet won't re-trigger until you close the tab.
• "Add to allowlist" — permanent. The snippet (or any snippet containing it, at least 8 chars long) is silently ignored from then on. You can review and remove allowlist entries in Options → Allowlist.

If you're seeing the same false positive often and think the regex is wrong, please open an issue with the snippet redacted — improving the detector helps every user.

Open the PasteGuard popup from the Chrome toolbar → click Upgrade to Pro → choose Monthly ($4.99/mo) or Annual ($39/yr). Checkout opens on ExtensionPay (which uses Stripe). Pro unlocks automatically once payment completes.

If you've already paid on another browser or device, click Sign in in the popup and use the email you paid with — Pro status will sync.

Open the popup → click Manage Pro → opens the ExtensionPay customer portal where you can cancel. Your Pro features stay active through the end of the period you've paid for. No prorated refunds.

No. Every detection happens in your browser. PasteGuard makes zero outbound network requests of its own. You can verify this yourself: open DevTools → Network tab on any AI chat site, paste a fake AWS key like AKIAIOSFODNN7EXAMPLE, watch the modal appear, and notice that the Network panel stays empty.

See the full privacy policy for details.

The semantic layer uses Chrome's built-in Gemini Nano model, which has hardware requirements:

• ≥22 GB free disk space
• >4 GB VRAM or ≥16 GB RAM with a 4-core CPU
• Windows 10/11, macOS 13+, Linux, or ChromeOS (no Android/iOS)
• Chrome 138 or later

If your hardware qualifies, semantic detection should be available. First use triggers a ~2-4 GB model download. The regex layer still works on devices that can't run the model.

Open the popup → click Pause for 1 hour. The extension will resume on its own. Useful if you're working through a stream of pastes that you know are safe and don't need the modal interrupting.

Yes. Open Options → Custom rules → Add rule. Each rule has a name, a regex pattern, a severity level, and a flag. The live tester shows whether your pattern matches sample text before you save.

The free tier allows up to 5 custom rules. Pro is unlimited. Patterns that are likely to backtrack catastrophically (nested quantifiers, alternation with outer quantifier, etc.) are rejected at save time.

In Chrome: right-click the PasteGuard toolbar icon → Remove from Chrome… Confirms. Or go to chrome://extensions/, find PasteGuard, click Remove.

Uninstalling automatically clears all locally-stored settings, audit log entries, and Pro status. Nothing is left behind on your machine or on any server (there are no servers).

The detection rules (regex patterns + semantic prompts) are open-source at github.com/Matteo-Coder2/pasteguard-rules. You can audit exactly what the extension looks for.

The extension code itself (popup, options, content script glue) is currently closed-source. If you want a specific assurance about a behavior, ask via support — happy to share specifics.

Please email support@pasteguard.io with [SECURITY] in the subject line. Avoid public disclosure (GitHub issues, social media) until we've had a chance to address it. We'll credit you on the fix announcement if you'd like.