Source code of the paper "Why Are My Prompts Leaked? Unraveling Prompt Extraction Threats in Customized Large Language Models"

These are my personal copies of the leaked system prompts for all of the popular LLMs

Security-hardened Claude Code skill with 3 defense layers against prompt injection, data exfiltration, and credential leaks. Companion repo for the YouTube video: 1 Prompt Hacked My Claude Code Agent

This is my first python project in the cybersecurity domain which is basically a keylogger - that is it records the key prompts typed by the vulnerable users and sends the typed text via mail to the developer (or the mail specified) probability leaking out passwords and vital information