secureonelabs/garak

garak is a free, open-source tool for testing and assessing the security of large language models (LLMs) by probing for vulnerabilities like prompt injection, data leakage, and toxic output. It supports a wide range of LLM providers and model types, allowing users to systematically evaluate model robustness against various attack scenarios. This tool is primarily designed for AI researchers, security professionals, and developers who want to understand and mitigate potential risks in their LLM applications.