Browse

Threat-Modeling

The Envelope: Contracts and Trust Boundaries for Agentic AI

Nov 5, 2026 12:00AM
19 min read
4007 words
The Envelope: Contracts and Trust Boundaries for Agentic AI
An office-envelope analogy for keeping AI planning, policy authorization, execution, and audit evidence separate.
software engineering artificial intelligence security
ai-security agentic-ai authorization contracts governance threat-modeling