AI Attack Guardrail

This is a prompt security skill created by The Gen AI that helps AI assistants and coding agents detect suspicious, malicious, or unsafe instructions inside files, code, skill packages, and prompts before you run them.

To use it, drag the skill folder or ZIP into Claude, ChatGPT, ChatGLM, a coding agent, or plug it into your existing AI workflow or internal solution.

It will scan for risks like hidden instructions, credential theft, unsafe system access, suspicious commands, and prompt injection attempts, then warn you before you install, execute, or trust anything.