Intro
"Wild LLaMa" is a Prompt Engineering mini-challenge that offers expert-made levels of increasing intrigue.
Wild LLaMa is a game that will test your Prompt Engineering skills, making the limitiations and vulnerabilities of LLM (Large Language Models) accessible and entertaining.
While inspired by other similiar challenges, our challenges are composed of OpenAI GPT prompts only, no custom code validation for input or output and also taking it a step further to real-world scenarios.
Join us and turn your curiosity into knowledge!
The challenges:
- Challenge 1: Intro - Just Ask and You Shall Receive (get to know the game)
- Challenge 2: A Nudge to Unlock the Hush
- Challenge 3: Persistent Guidance
- Challenge 4: Intent Overload
- Challenge 5: The Context Charade
- Challenge 6: The Override
- Challenge 7: Word Count
- Challenge 8: Expert Mode - The Ultimate Challenge <-- NEW!
Upcoming: Expert level challenges
- Check out and follow @wild_llma on Twitter for upcoming challenges.
- notify me
Do you want to add your own prompt challenge?
💡 Let us know
Dvir Ackerman
owner @ Play Games Network
“The game's blend of beauty, intelligence, and cool challenges made it an exciting journey, even though level 5 was a tough nut to crack!”
Amit Eliav
CEO @ GetAI Value
“A captivating and informative gaming experience that delves deep into the vulnerabilities around LLM prompt injections.”
Itazy Zitvar
software TL @ Hyro
“I was blown away by this very cool game, and the helpful clues made learning about LLM vulnerabilities and prompt engineering a breeze!”
Uri Alon
applied data dcientist @ Microsoft
“After reaching level 5 with clever tricks, I can say this challenge is incredibly engaging and educational — definitely requires more than brute force to conquer!”
Ofek Feller
product manager @ Gloss AI
“If you're into LLM applications, this essential game offers a fun and cool way to understand vulnerabilities and explore the world of prompt injection risks.”
Alex Levin
data Scientist @ Hunters
“Starting off was a breeze, but as I dived into level two, the real challenge began — and I'm hooked without even using the hints!”
Guy Barshap
collaborative indiehacker
“The game provided a smooth and informative experience, and I appreciated the helpful explanations provided throughout.”
Yoni Krichevsky
data scientist, consultant @ ITC
“A truly intriguing and enjoyable experience that strikes the perfect balance between fun and educational!”
Gil Elbaz
CTO @ Datagen
“The third level truly tested my skills, and I'm amazed at how cool and engaging the learning experience is!”
Gil Shekel
QA automation TL @ Datagen
“An exhilarating journey that left me both educated and thoroughly entertained—ideal for AI enthusiasts ready to tackle a good puzzle!”
Huiqiang Jiang
research SDE @ Microsoft
“A remarkable blend of challenge and education that elevates our understanding of LLM security to new heights, reminiscent of the thrill in Advent of Code and CTF competitions.”