AI Agent Architecture: The Trust Boundary Model

AI Agent Memory vs State: What Should Be Remembered, Stored, or Recomputed?

Tool Output Is Not Instruction: A Core Rule for Secure AI Agents

Secure Architecture for AI Agents That Read Email, Documents, and Webpages

A practical series on building AI agents that are secure by design — covering trust boundaries, prompt injection defenses, scoped tool access, approval gates, and audit architecture. Each post stands alone; together they form a complete playbook.

AI Agent Prompt Injection Risk Scorecard

Human-in-the-Loop AI Agents: Where Approval Gates Actually Matter

Designing Production-Grade AI Agents: Permissions, Tools, Logs, and Rollbacks

Building AI Agents That Can Use Tools Without Owning Secrets

AI Agent Audit Logs: What Enterprises Need to Capture

AI Agent Runtime Control: Why Prompt-Level Guardrails Are Not Enough

RAG vs Agent Memory vs Workflow State

AI Agents in Regulated Enterprises: Access, Approval, Audit, and Deployment Constraints

Designing Secure AI Agents