GitHub repository raaihank/llm-sentinel provides a privacy-first proxy server that automatically detects and masks sensitive data before it reaches AI models. It offers PII detection, prompt injection protection, a real-time dashboard, multi-provider support, and zero configuration deployment.