The security library / 01—05
Understand the system.
Secure the boundary.
Field notes on AI, applications, infrastructure, cryptography, and systems. Concise explanations. Visible assumptions. Concrete failure cases.
Five courses, each built the same way: the idea first, then the picture, the mechanism, a program you can run, and the place where it breaks.
Choose a subject
Follow a path. Go deep.
- 0122 lessons
AI security engineering
How AI systems are attacked and how they are defended, from the first threat model to a running agent.
- 0221 lessons
Application security
Identity, permissions, untrusted input, and the mistakes that turn an API into an attack surface.
- 0319 lessons
Infrastructure security
Workload identity, secrets, isolation, and the trust decisions underneath a running service.
- 0420 lessons
Cryptography
Keys, signatures, and encryption, connected to the assumptions that make them useful.
- 0522 lessons
Systems design
Invariants, failure modes, concurrency, and the tradeoffs behind systems that keep their promises.
A good place to begin
Draw the system.
Then draw the trust.
Start with AI security engineering: why it is different, how attacks work, and which defenses hold when the model is fooled.
Open the course