YMS360 AI Prompt Shield

Review yacht-sensitive information locally before inserting prompts into ChatGPT or Claude.

YMS360 AI Prompt Shield helps yacht crew keep sensitive information out of ChatGPT and Claude. Use ChatGPT or Claude as usual; the Shield checks your message right in the message box, on your device. It's free, and needs no account and no API key. HOW IT WORKS • A small 🛡 indicator shows whether your message contains recognized sensitive details, such as vessel names and IMO/MMSI numbers, owner and guest movements, internal IP addresses, passwords and keys, emails, phone numbers, and bank details. • Press Send with something sensitive in it and sending is paused. Choose "Replace all" to swap the details for placeholders like [VESSEL_1], or "Send anyway". Then press Send yourself; the Shield never sends for you. • Paste sensitive text and the Shield checks it before the website receives it, offering a cleaned version. • For maximum privacy, open the side panel and draft there. It also shows whether the current tab is protected. PRIVACY • All checking happens on your device. Nothing is sent to YMS360. No analytics, no tracking, nothing stored. • Text you type in ChatGPT or Claude is visible to that website as you type. The Shield checks it before it is sent. • Works only on chatgpt.com and claude.ai. LIMITATIONS This tool helps reduce accidental disclosure; it cannot identify every sensitive detail. It does not check attachments, images, voice input, earlier chat messages, or other websites. If the 🛡 indicator is not shown, that message box is not being checked. YMS360 is not affiliated with OpenAI or Anthropic.