Collected sources and patterns will appear here. Add from search or the patterns library.
Prompt -> List<AdversarialPrompt>
Synthesize variations of test prompts using semantic manipulation operators to test LLM safety guardrails.
Problem it solves
Manual evaluation of LLM robustness is insufficient to cover the diversity of jailbreak techniques.
Consumes
Emits
The real projects this mechanism was found in. Attribution is the point — this is how the best teams actually do it.