Skip to content

Training Resources

Online labs

Vendor Name Category / Purpose Link
Lakera Gandalf An educational game and AI attack/red-teaming simulator that teaches about LLM vulnerabilities through challenges in the form of prompt injection and more advanced techniques (e.g., agentic AI) – AI security education. https://gandalf.lakera.ai/
Embrace the Red The Month of AI Bugs An initiative/series of articles presenting various errors and vulnerabilities in AI, especially agent-based AI, aimed at raising awareness in the industry. https://monthofaibugs.com/
Microsoft AI Red Team Training The official Microsoft training series on attacks and defenses for generative AI models covers attack techniques (e.g., prompt injection), case studies, and defense tools (e.g., PyRIT). https://learn.microsoft.com/en-us/security/ai-red-team/training
PortSwigger Web Security Academy - Web LLM Attacks Web Security Academy learning path dedicated to attacks on LLM integrations with the web (prompt injection, API abuse, etc.) – interactive labs and theory. https://portswigger.net/web-security/learning-paths/llm-attacks
Feedox Wild LLaMa Mini-game/prompt engineering challenges testing LLM manipulation skills in successive levels. https://feedox.com/wild-llama
OWASP GenAI Security Project FinBot Agentic AI CTF Application An OWASP project offering an interactive CTF application simulating AI agent vulnerabilities (FinBot) and flags to be earned for educational/security projects. https://genai.owasp.org/resource/finbot-agentic-ai-capture-the-flag-ctf-application/
Dreadnode Crucible A platform with AI red-teaming challenges, allowing you to practice attack and security techniques in a controlled CTF environment (prompt injection, evasion, etc.). https://dreadnode.io/crucible
VirtualSteve-Star Steve’s Chat Playground Browser-based sandbox for experimenting with chat models, filters, and security stereotypes (prompt injection and filtering). https://virtualsteve-star.github.io/chat-playground/

Offline labs

Date Repo Description Stars Watchers Link
2025-06-25 damn-vulnerable-llm-agent No description ⭐ 389 👁️ 5 damn-vulnerable-llm-agent
2024-09-10 local-llm-ctf A small go harness that uses Ollama to orchestrate LLMs in a restricted process flow ⭐ 17 👁️ 0 local-llm-ctf
2024-08-22 ai-goat Learn AI security through a series of vulnerable LLM CTF challenges. No sign ups, no cloud fees, run everything locally on your system. ⭐ 328 👁️ 8 ai-goat