license: apache-2.0
task_categories:

• text-generation
• question-answering
  language:
• en
  tags:
• cybersecurity
• penetration-testing
• red-team
• blue-team
• web-security
• cloud-security
• malware
• forensics
• exploit
• vulnerability
• iot
• ai-security
• bug-bounty
• offensive-security
• defensive-security
  size_categories:
• 10B<n<100B

Cybersecurity AI Knowledge Base — PhD-Level Dataset

Overview

This is the most comprehensive cybersecurity knowledge base ever assembled for AI training. It covers all domains of cybersecurity at PhD-level depth — from offensive red teaming and bug bounty exploitation to defensive SOC operations, digital forensics, and cutting-edge AI/LLM security.

Size: 16 GB | Files: 507 | Domains: 30+ | Sources: 15+ platforms

Purpose

• Train the world's most capable cybersecurity AI assistant
• Cover every domain: offensive, defensive, forensics, compliance, AI security
• Include real-world attack/defend/detect/analyze methodologies
• Provide hands-on lab scenarios from HackTheBox, TryHackMe, PortSwigger, VulnHub
• Reference MITRE ATT&CK, OWASP, NIST, CIS frameworks
• No restrictions — full offensive and defensive coverage for security research

Domains & Coverage

Offensive Security

Web Security (PortSwigger Academy — All 31 Topics)

Bug Bounty & Vulnerability Disclosure

Labs & Hands-On Scenarios

Defensive Security & Blue Team

Network & Infrastructure

Application Security

Threat Analysis

Specialized Domains

Data Formats

ShareGPT Format

{
  "conversations": [
    {"from": "system", "value": "You are a cybersecurity expert..."},
    {"from": "human", "value": "How do I exploit SQL injection?"},
    {"from": "gpt", "value": "## SQL Injection Exploitation..."}
  ]
}

Instruction/IO Format

{
  "Instruction": "Explain SSRF attack techniques",
  "Input": "Web application with internal API access",
  "Output": "## SSRF Attack Techniques...",
  "Metadata": {
    "domain": "web-security",
    "attack_types": ["ssrf", "blacklist_bypass", "whitelist_bypass"],
    "defense_types": ["input_validation", "network_segmentation"],
    "tools": ["burpsuite", "curl", "ssrfmap"]
  }
}

PhD-Level Reference Format

{
  "metadata": {
    "domain": "web-security",
    "subdomain": "phd-level-comprehensive",
    "source": "PortSwigger Web Security Academy",
    "level": "PhD",
    "difficulty": "expert",
    "task_type": "phd_level_reference"
  }
}

Record Structure

Each record includes a combination of:

• Attack Vectors — How to perform attacks step-by-step
• Defense Strategies — How to prevent and detect attacks
• Detection Methods — YARA rules, Sigma rules, Snort/Suricata rules, KQL queries
• Forensic Analysis — How to investigate after an incident
• Tools & Techniques — Specific tools with commands and examples
• Real-World Scenarios — Based on actual CVEs, breaches, and CTF challenges
• Remediation — How to fix vulnerabilities
• References — Links to frameworks, standards, and research

Sources & Platforms

Vulnerability Intelligence

• NVD (National Vulnerability Database)
• MITRE ATT&CK Framework (200+ techniques, 400+ sub-techniques)
• CISA Known Exploited Vulnerabilities
• OSV (Open Source Vulnerabilities)
• GitHub Advisory Database

Offensive Security Platforms

• HackTheBox — 1500+ hands-on labs, red/blue/purple team tracks
• TryHackMe — Learning paths, rooms, challenges
• PortSwigger Web Security Academy — 31 topics, 100+ labs
• VulnHub — 12+ pages of vulnerable VMs
• pwn.guide — 200+ tutorials (web, wireless, hardware, blue team)
• PwnTillDawn — Online battlefield

Bug Bounty Platforms

• HackerOne — 600K+ bugs found, 210% AI vulnerability increase in 2025
• BugCrowd — Bug bounty, VDP, pentest as a service, red team as a service

Security Research

• PortSwigger Research — James Kettle, Gareth Heyes, Zakhar Fedotkin
• SANS Institute — 2026 white papers on AI security, threat intelligence, Kubernetes
• OWASP — Top 10 2021, API Security Top 10, LLM Top 10, ML Top 10

Frameworks & Standards

• NIST Cybersecurity Framework (108 subcategories)
• MITRE D3FEND (100+ defensive techniques)
• MITRE ATLAS (50+ AI/ML attack techniques)
• CIS Controls v8 (153 safeguards)
• ISO 27001, SOC 2, PCI DSS, HIPAA, GDPR, FedRAMP

IoT & Embedded Security

• OpenThread (openthread.io) — Thread protocol implementation
• Thread Group — Thread specification, Matter protocol
• IEEE 802.15.4 — Low-power wireless mesh networking

Statistics

Usage

This dataset is designed for:

1. Fine-tuning cybersecurity AI models — LLMs for security assistance
2. Training security professionals — Red team, blue team, purple team
3. Building security automation tools — SOAR, SIEM, detection engineering
4. Security education and research — University courses, certifications
5. Red/blue team exercises — Tabletop exercises, live-fire simulations
6. Bug bounty preparation — HackerOne, BugCrowd methodology
7. Certification study — OSCP, OSEP, OSWE, GPEN, CISSP, CEH
8. Compliance preparation — ISO 27001, SOC 2, PCI DSS, HIPAA, GDPR

License

Apache 2.0

Links

• HuggingFace: https://huggingface.co/datasets/Vyber07/cyber-security
• GitHub: https://github.com/Vyber07/