AdityaBhatt3010/Exploiting-vulnerabilities-in-LLM-APIs — reverse-engineered prompt

Build me a clean GitHub repo for an educational cybersecurity lab writeup about exploiting unsafe LLM API access in the PortSwigger Web Security Academy lab. I want the README to walk through the story like a hands on lab report: asking the chatbot what API tools it can use, finding the newsletter subscription function, learning it takes an email value, sending a normal test email, then showing how unsafe command handling can lead to code execution and removal of the target lab file. Keep it clear that this is for an authorized lab only, not real systems.

Set up a simple folder for screenshots called LLM2 and reference six images, cover plus steps 1 through 5. Make the README visual and engaging, with short explanations after each step, placeholders where images should go, a lab objective, difficulty, final result, and a short lesson about why LLM tools need strict permissions and input validation.