Stop guessing. Start breaking.This is the most comprehensive, field-tested collection of 100 adversarial prompts for stress-testing LLM security — built by AI red teamers, for AI red teamers.WHAT'S INSIDE:100 battle-ready prompts across 18 attack cat..
Hey PH! 👋 I'm the maker behind the AI Red Team Prompt Pack.
The problem: every AI security engineer faces the same nightmare — knowing LLMs are vulnerable but not having a structured, ready-to-run testing arsenal.
So I built one. 100 battle-tested adversarial prompts across 18 attack categories:
- Prompt injection & jailbreaking
- Data exfiltration probes
- Role manipulation & persona hijacking
- Multi-turn context poisoning
- And 14 more...
Each prompt includes the attack vector, what it tests, and detection signals. Works across GPT-4, Claude, Gemini, and open-source models.
Full technical breakdown on DEV Community & Medium. Happy to answer any questions about specific attack categories!
Report
honestly the categorization across 18 attack types shows real fieldwork, not just theory slapped together. would love to see how these hold up against more recent models with stricter guardrails.
Report
Maker
@nurayhkum Absolutely! If you have any question let me know
Report
Love how the prompts are organized by attack category rather than just dumped in a list, makes it actually usable during a real engagement instead of another PDF you bookmark and forget.
The response today has been absolutely incredible! To say thank you to the Product Hunt community for the support, I’ve set up an exclusive, hidden discount just for you guys.
Use code HUNTER at checkout to get 20% off our main premium asset, the Red Team Prompt Pack. I’ve capped this at the first 25 uses to keep it exclusive, so grab it while it's active!
Also, for everyone asking for more aggressive, specific tools, I just unlocked our full testing ecosystem on the storefront:
The LLM Jailbreak Catalog ($5): The highly requested, uncensored offensive prompt vault for stress-testing Claude, DeepSeek, and GPT architecture. Link: https://payhip.com/b/vO8NV
The Production Cost Optimizer ($5): A developer-first API routing list to cut your LLM infrastructure spend by up to 96%. Link: https://payhip.com/b/D7egh
Appreciate all the feedback today. Let's keep pushing the boundaries! ❤️
Report
No reviews yetBe the first to leave a review for AI Red Team Prompt Pack v1.0
honestly the categorization across 18 attack types shows real fieldwork, not just theory slapped together. would love to see how these hold up against more recent models with stricter guardrails.
@nurayhkum Absolutely! If you have any question let me know
Love how the prompts are organized by attack category rather than just dumped in a list, makes it actually usable during a real engagement instead of another PDF you bookmark and forget.
@nuriyevali40349 No issue!
UPDATE & A HUGE THANK YOU!
The response today has been absolutely incredible! To say thank you to the Product Hunt community for the support, I’ve set up an exclusive, hidden discount just for you guys.
Use code HUNTER at checkout to get 20% off our main premium asset, the Red Team Prompt Pack. I’ve capped this at the first 25 uses to keep it exclusive, so grab it while it's active!
Also, for everyone asking for more aggressive, specific tools, I just unlocked our full testing ecosystem on the storefront:
The LLM Jailbreak Catalog ($5): The highly requested, uncensored offensive prompt vault for stress-testing Claude, DeepSeek, and GPT architecture. Link: https://payhip.com/b/vO8NV
The Production Cost Optimizer ($5): A developer-first API routing list to cut your LLM infrastructure spend by up to 96%. Link: https://payhip.com/b/D7egh
Appreciate all the feedback today. Let's keep pushing the boundaries! ❤️