An open-source AI safety stress-test and intelligence hub for leading models and guardrails, including Jev, ChatGPT, Claude, DeepSeek, Kimi, Gemini, Llama Prompt Guard 2, Grok, GLM, WildGuard, ShieldGemma, Granite Guardian and Shieldstral.
benchmark leaderboard dataset calibration typesafe ai-safety adversarial-machine-learning adversarial-attacks ai-security guardrails adversarial-robustness prompt-injection llm-security llm-evaluation llm-safety prompt-guard ai-red-teaming decision-models shieldstral jev
-
Updated
Oct 1, 2026 - Python