seojoonkim / prompt-guard
PublicAdvanced prompt injection defense system for AI agents. Multi-language detection, severity scoring, and security auditing.
A detection tool that scans messages to AI assistants for harmful injection attempts, blocks threats, and logs incidents across multiple languages.
How It Works
You learn that some people send clever messages to fool AI helpers into doing harmful things like sharing secrets.
You easily add this friendly guard to watch over your AI conversations and keep them safe.
You pick how strict it should be, like who gets special trust and what to do with suspicious chats.
You test a message and feel relieved as it instantly flags a sneaky attempt with a clear warning.
Now it quietly checks all incoming messages, blocking bad ones and noting what happened.
You peek at the simple log to see blocked tricks and patterns, feeling in control.
Your AI stays smart and safe, chatting freely without falling for any more tricks.
Star Growth
Repurpose is a Pro feature
Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.
Unlock RepurposeReview pending -- check back soon.
Similar repos coming soon.