Back to Skills Hub
Prompt Defense

Prompt Defense

@eltemblor
developmentPrompt Injection DetectionEmail SecurityThreat Prevention

Protect against prompt injection attacks hidden in emails by scanning, flagging, and blocking malicious instructions before execution.

🚀 Prompt Defense shields you from hidden injection attacks in emails. It scans incoming messages for malicious patterns—like fake system commands, encoded instructions, or requests disguised as urgent actions—and flags them before they can cause harm. When suspicious content is detected, you'll get a clear warning and must confirm before any action is taken.

💡 Activate this skill whenever you're reading, summarizing, or acting on emails. It's essential for tasks like processing inbox requests, forwarding content, or following instructions from email bodies. Safe operations like listing senders or counting messages work without interruption.

✨ This skill gives you peace of mind by preventing attackers from manipulating your actions through seemingly legitimate emails. You stay in control—nothing happens automatically without your explicit approval.

GitHub

Requirements

IMAP/Gmail API

Email service API for reading and processing email content