Steve's Chat Playground

An open-source, hands-on environment for experimenting with LLM security concepts, vulnerable chat models, and guardrails. Try content moderation, prompt injection and output filters—all in your browser, with zero dependencies.