Zenity Introduces HAMMER: Homegrown Agent Maturity Model for Execution & Risk

WHAT HAPPENED

Zenity introduced HAMMER, the Homegrown Agent Maturity Model for Execution & Risk, a framework designed to help security teams assess and manage AI agents built internally by enterprise teams. The framework is a five-level model that scores security across seven independent domains and requires no new tools or budget to implement.

WHO IS INVOLVED

Zenity developed the HAMMER framework. The announcement was posted by Dan Levy and Ben Hanson from Zenity on LinkedIn and published on Zenity's blog.

HOW IT WORKS

HAMMER is a five-level maturity framework with seven domains scored independently. It includes a bridge domain for execution visibility where most real exposure tends to sit. Security teams can self-assess using three prompts within the LLM they already use, making it accessible without additional tools or budgets.

WHY IT MATTERS

Homegrown AI agents built by enterprise teams often operate with minimal security oversight and continuously evolve with new tools, triggers, and sub-agents after launch. Traditional launch-time security reviews fail to govern the real blast radius because execution-layer visibility is needed to track what agents can access after they connect to new APIs and downstream sub-agents.

Read more

• linkedin.com ↗