Pick a depth. Each prompt opens in your AI pre-loaded with the lesson. Click a row to preview the prompt.
Security work is fundamentally about disagreement: every attacker has read the same docs as you, and they're looking for the cases the docs didn't cover. The defender's mindset is the discipline of imagining what an adversary would do with your system before they do it, then writing code (and policies) so that the adversary's best move is still a loss.
Same login form, two readings: a developer sees 'username and password'; a defender sees 'two input fields whose contents I can't trust, an auth function with timing characteristics, and an error surface that may leak account existence'.
' OR 1=1 -- in the email field'), not vague ('inject SQL').