Mask secrets before
you hit send.
Offsend catches secrets and personal data in your prompts and masks them automatically — before they ever reach ChatGPT, Claude, Gemini, Grok, Perplexity, or DeepSeek.
No cloud scanningNo prompt content sent to serversNo account required
Scan locally.
Send safely.
It all runs locally as you type. Offsend flags risky values, masks them the way you choose, and brings the originals back whenever you need them.
Review
See every sensitive match before send and uncheck any value you want to keep in the prompt.
Learn moreDetectors
Toggle built-in scanners for emails, API keys, tokens, private keys, and other values that leak into AI chats.
Learn moreMode
Choose Warn, Auto-mask, or Block when sensitive data appears — review first or stop the send entirely.
Learn moreMasking
Replace secrets with stable placeholders and restore originals locally for a window you control.
Learn moreCustom rules
Add JavaScript regex patterns for company-specific values — matched alongside built-in detectors.
Learn moreReview every match
before you send.
Before you send, Offsend lists every sensitive match in your prompt. Uncheck an item to leave the original text — only checked values are replaced with placeholders.
Install for Chrome
Pick what
gets scanned.
Built-in detectors cover the data that most often leaks into AI chats. Toggle each type on or off in the extension settings.
Warn, mask,
or block.
Pick how Offsend reacts when it finds sensitive data in a prompt. Review findings first, mask quietly in the background, or stop the send until values are masked.
Default is Warn: you see what was found and choose Mask & send, Send anyway, or Cancel.
Placeholders you
can restore.
Sensitive values become placeholders like {{API_KEY_1}} so your prompt still reads clearly to the AI. Encrypted mappings are kept locally for a window you choose.
Set the restore window in minutes — from a quick 15-minute session to up to 24 hours. When it expires, mappings are deleted and placeholders stay in place.
Your regex.
Your rules.
Built-in detectors cover common leaks, but every team has its own formats. Add JavaScript regex patterns that run alongside them — and toggle Custom rules under Detectors when you want them active.
Enter patterns without /.../ delimiters — for example, internal IDs, ticket prefixes, or proprietary token formats your org uses every day.
Send safer
prompts today.
Add Offsend to Chrome or Firefox and mask secrets and personal data automatically — before they reach ChatGPT, Claude, or Gemini.