LLM Canary

LLM Canary is a small open framework for testing models against security vulnerabilities, organized around the OWASP LLM Top 10: prompt injection, insecure output handling, data leakage, and so on. You run its test suites against a model and get a scorecard per category. It came out of a security research group and is lighter weight than PyRIT or garak.

Fine as a quick benchmark; less active than the larger tools.