GitHub falsely banned me for open-source defensive security code.

<p>My primary GitHub account (Shlok-Parekh09) was caught in an automated ban a month ago. The trigger was a hackathon project called CodeGuard—a real-time IPS layer for LLM-generated code. My training datasets contained inert signatures (fake API keys, SQLi strings, rm -rf) strictly to train the model to detect and block them. The automated system flagged this defensive security research as actual

Read the source →