RedVector
Adversarial LLM testing framework - internally, BurpSuite for AI agents.
- A platform for poking holes in LLMs before they go live: prompt injections, jailbreaks, adversarial edge cases. It turns what breaks into a score a team can actually act on.
- FastAPI
- LiteLLM
- React