Blog
Technical articles on AI agent security, runtime monitoring, and policy-based controls.
Prompt Injection: The Practical Threat Model
Understanding how prompt injection attacks work against tool-using AI agents and what runtime controls can prevent them.
Why Antivirus Doesn't Stop Tool-Using Agents
Traditional antivirus products aren't designed for the threat model of autonomous AI agents. Here's what's different.
Local-First AI Security: What It Means
Why AI agent security should run on your device, not in the cloud. The case for local-first runtime monitoring.
Policy-Based Controls for Agents
How policy rules let you define exactly what AI agents can and cannot do on your system.
Audit Logs for AI Tool Use
Why every AI agent action should be logged, and how audit trails enable accountability and debugging.
Rogue Agent Scenarios (Safe Examples)
Real-world examples of how AI agents can behave dangerously, and what runtime controls would catch them.