fabraix/playground
A live environment to stress-test AI agent defenses through adversarial play 🧠
30 best points · 19 comments · 2 Hacker News posts · 76 GitHub stars · View on GitHub
Hacker News discussions
- Show HN: Open-source playground to red-team AI agents with exploits published (30 points, 13 comments, by zachdotai, 2026-03-15)
- AI agents are easy to break (4 points, 6 comments, by zachdotai, 2026-02-11)