Guides
Practical, no-fluff writing on AI agent security — how the attacks work and how to defend against them.
What Is Prompt Injection? A Plain-Language Guide for Agent Owners
Direct vs. indirect injection, why agent tools make it dangerous, and the layered defenses that hold.
2026-10-11 · 9 min read
How to Stop Your AI Agent From Leaking Secrets
How data exfiltration happens through agents, why system prompts don't stop it, and the boundary defenses that do.
2026-10-11 · 9 min read
Canary Tokens for AI Agents: Early-Warning Tripwires
Fake secrets planted where agents work — when one fires, you know something went rogue. Deployment guide.
2026-10-11 · 9 min read
Policy Enforcement for Autonomous Agents: Approvals, Quarantine, and Kill Switches
Stop trusting the model and start policing actions: allow/approve/deny gates, quarantine, and kill switches.
2026-10-11 · 9 min read
Agent Permission Scoping: Give Your AI Agent the Smallest Possible Keys
Excessive agency (OWASP LLM06) turns a confused agent into a dangerous one. The least-privilege playbook: scoped tokens, read/write splits, network allowlists, spending caps.
2026-10-13 · 9 min read
📋 Free resources
Prompt Injection Defense Cheat Sheet — the 7 attack families, 5 defenses, 60-second self-check. Free one-page PDF, no signup.
2026-10-13 · cheat sheet