
FreakLabs
Hands-on AI security: prompt injection & agent attacks
About FreakLabs
FreakLabs is a hands-on AI security training platform built by KageX, an offensive AI security startup in Noida, India. Learners attack real LLMs, AI agents and RAG systems in private sandboxes across 20 attack labs and a 9-level, 54-lesson curriculum covering prompt injection, jailbreaks, data leakage, insecure agent tool use and poisoned retrieval. Two labs are free; full access is $29.99/month. Built for security engineers, pentesters, AI developers and students.
What FreakLabs does
AINave analysisUsers practise attacking live AI models, RAG pipelines and agents in private sandboxes to learn AI security failures.
- Best for
- Security engineersPenetration testersAI developersStudents
- Use it to
- Learning prompt injection and jailbreak techniques
- Practising indirect injection and RAG poisoning
- Testing agent tool misuse in sandboxed environments
- Completing guided AI security lessons alongside labs
- Team and college security training
Show you're on AINave
Add the badge to FreakLabs's site. It links back to this page.