README.md

metadata

license: apache-2.0
task_categories:
  - text-classification
language:
  - en
tags:
  - security
  - LLM-safety
  - prompt-injection
  - jailbreak
  - adversarial-ai
  - ai-security
  - llm-security
  - red-team
  - prompt-defense
  - ai-firewall
  - cybersecurity
  - instruction-override
  - system-prompt-protection
  - dataset
  - ai-defense
size_categories:
  - 100K<n<1M

The RedLockX Dataset is a large-scale curated security dataset designed for evaluating and training AI systems against adversarial threats such as prompt injection, jailbreak attempts, system prompt leakage, and LLM manipulation attacks.

It contains structured real-world and synthetic attack patterns used in modern AI red-teaming.

📌 Dataset Overview

✔ 109,000+ labeled adversarial & safe samples
✔ Multi-category threat classification system
✔ OWASP LLM Top 10 mapping included
✔ Severity scoring (0–10 scale)
✔ Risk scoring (0–100 business impact model)
✔ Suitable for fine-tuning, evaluation, and benchmarking

🚀 How to Use

You can load this dataset using the Hugging Face datasets library.

• Use for prompt injection detection models
• Train LLM guardrails and safety classifiers
• Benchmark adversarial robustness
• Red-team AI systems before deployment

⚖️ License

This dataset is released under the Apache 2.0 License.
It is intended strictly for research in AI safety, adversarial robustness, and security evaluation.