Think like the adversary.
Field manuals and tools for offensive security — from recon-to-red-team fundamentals to breaking and defending AI. Learn the attack. Prove your defense.
Prove where your AI actually stands.
The AI Audit is a free, browser-based self-assessment. Answer honestly and it returns a readiness score, a prioritized gap list, and a report you can hand to an engineer, an executive, or an auditor.
- Classify your app under the EU AI Act — role, reach, risk tier.
- Score it against all ten OWASP LLM Top 10 controls.
- Get a gap register and a deadline-anchored remediation plan.
Free · runs entirely in your browser · nothing leaves your device · not legal advice.
until the EU AI Act's transparency rules and GPAI enforcement apply — 2 Aug 2026.
Two field manuals. One operator's mindset.
Buy either bundle on its own, or take the whole shelf. Every purchase is a one-time payment, an instant download, and free updates to its edition.
The bundle
The original — recon to red team.
- 122-page field manual, 14 chapters
- 120-page Lab Workbook — 100+ labs
- Appendices A–E + free edition updates
The bundle
Break the machine — LLM & agentic apps.
- 57-page AI red-team guide, 10 chapters
- 103-page Lab Workbook of hands-on labs
- Mapped to OWASP LLM Top 10 & MITRE ATLAS
The whole shelf
Both field manuals + both lab workbooks.
- The Adversary's Mind + its Lab Workbook
- Breaking AI + its Lab Workbook
- Everything above — $88 value
Only need one book? The Adversary's Mind or Breaking AI is $39 and $29 respectively on its own page. All sales are final once a file is downloaded; if you haven't downloaded yet, request a full refund within 7 days.
Not a command dump. A way of thinking — the attacker's.
The Adversary's Mind is a small, independent shop making offensive-security field manuals for people who'd rather understand a system than memorize a tool. The same operator's mindset runs through everything here — the classic recon-to-red-team manual, the AI red-teaming line, and the tools that let you prove your own systems hold up. Everything is written for systems you own or are explicitly authorized in writing to test.