Analysis · curated 13 Aug 2026
The Agent Baseline: 35 controls, but where should you start?
First reported snyk.io
Coverage timeline
Single-source analysis — first reported, latest, and curated coincide.
Why it matters
The Agent Baseline offers defenders a shared vocabulary and control framework for securing enterprise AI agents, helping them distinguish overlapping vendor claims and spot gaps in agent governance coverage.
Snyk, alongside Docker and Keycard, published the Agent Baseline (agentbaseline.org), an open reference architecture defining six security outcomes and 35 controls for running AI agents at enterprise scale. This blog post reflects on the framework after a Black Hat panel, discussing how to prioritize where to start and how the six outcomes help practitioners map vendor claims and identify buy-versus-build gaps in agent security.