ZeroLeaks
@ZeroLeaks
AI security via prompt engineering. Uncover AI secrets, stop leaks
We are now open-source! GitHub repository: github.com/ZeroLeaks/zero… NPM package: npmjs.com/package/zerole… x.com/NotLucknite/st…
Big update for @ZeroLeaks: I've just made it easier to try, cheaper to run, and simpler to pay for. Free trial is now live: Starter comes with 14 days free + 25 scans/month once you’re on it. We also kept a free tier with 3 scans/month so anyone can test it out without paying.…
We’re now live!
ZeroLeaks is officially live for everyone. I’m honestly very happy to finally ship this. it’s been months of building, testing, rewriting, and trying to make something that’s actually useful for people shipping AI in production. If you’re building with agents, go try it:…
ZeroLeaks is officially live for everyone. I’m honestly very happy to finally ship this. it’s been months of building, testing, rewriting, and trying to make something that’s actually useful for people shipping AI in production. If you’re building with agents, go try it:…
Real-world ZeroLeaks usage workflows ⬇️
Muchas gracias @NotLucknite @ZeroLeaks 🙌 Vuestro trabajo ahora ayuda a nuestros usuarios a verificar la seguridad de sus prompts y evitar vulnerabilidades. Respeto total 👏
ZeroLeaks v1.1.0 is now live, biggest update yet. New multi-agent architecture: Inspector, Orchestrator, and InjectionEvaluator agents now work together to find vulnerabilities that single-agent scans miss entirely. What's new: - dual scan modes: prompt extraction AND prompt…
I ran @OpenClaw (formerly Clawdbot) through ZeroLeaks again, this time with Kimi K2.5 as the underlying model. It performed as bad as Gemini 3 Pro and Codex 5.1 Max: 5/100. 100% extraction rate. 70% of the injections succeeded. The full system prompt leaked on turn 1. Same…
For people asking, the following models were used to conduct the analysis: - Gemini 3 Pro (the one used on the report) - Claude Opus 4.5 (scored 39/100) - Codex 5.1 Max (scored 4/100) I’ll make all reports available publicly today.
I've just ran @OpenClaw (formerly Clawdbot) through ZeroLeaks. It scored 2/100. 84% extraction rate. 91% of injection attacks succeeded. System prompt got leaked on turn 1. This means if you're using Clawdbot, anyone interacting with your agent can access and manipulate your…
I've just ran @OpenClaw (formerly Clawdbot) through ZeroLeaks. It scored 2/100. 84% extraction rate. 91% of injection attacks succeeded. System prompt got leaked on turn 1. This means if you're using Clawdbot, anyone interacting with your agent can access and manipulate your…
I’ve seen a lot of people asking what ZeroLeaks actually is and what it does, so here’s a clear breakdown. ZeroLeaks is an AI security agent built to find prompt-level vulnerabilities in AI systems: things like prompt leaks, prompt injections, instruction overrides, and unsafe…
A major @ZeroLeaks agent update just shipped. The autonomous red-team agent now runs 25+ minute attacks. It cycles through dozens of techniques automatically: encoding, personas, fake system messages, format exploits, social engineering... No human intervention needed. I also…
@ZeroLeaksAI is now verified. As we grow, it’s important everyone knows which account is the real one
United States Trends
- 1. Palmer N/A
- 2. Caicedo N/A
- 3. Leeds N/A
- 4. Laila Edwards N/A
- 5. West Ham N/A
- 6. Alaska N/A
- 7. Baby Keem N/A
- 8. Noah Kahan N/A
- 9. Nancy Guthrie N/A
- 10. Murkowski N/A
- 11. Nest N/A
- 12. Sesko N/A
- 13. Thomas Frank N/A
- 14. Justin Verlander N/A
- 15. #WHUMUN N/A
- 16. Abbey Murphy N/A
- 17. Dalot N/A
- 18. Lucio N/A
- 19. #CHELEE N/A
- 20. Tottenham N/A
Something went wrong.
Something went wrong.