JFrog Ltd. (NASDAQ: FROG) announced the launch of its JFrog Agent Skills Registry, designed to provide governance and security for enterprise AI agents operating at scale. The platform was developed through integration with NVIDIA and supports the NVIDIA Agent Toolkit, including NVIDIA OpenShell.

The registry serves as a secure system for storing and managing Model Control Protocols (MCPs), models, agent skills, and agentic binary assets. JFrog Artifactory will function as a registry for AI models and agent skills with NVIDIA AI-Q Blueprint as part of the NVIDIA Agent Toolkit.

"AI agents are fundamentally reshaping how software is created and operated, but without a dedicated trust layer to enforce governance and secure workflows, they introduce significant enterprise risk," said Gal Marder, JFrog's Chief Strategy Officer.

The solution addresses security concerns in the AI agent ecosystem by providing scanning, verification, and governance capabilities. The platform automatically scans and verifies AI skills upon upload to detect vulnerabilities and malicious payloads before agents can access them.

Key features include a certified NVIDIA AI-Q Blueprint integration, native NVIDIA OpenShell integration, centralized agent system of record through JFrog AI Catalog, and policy-driven governance controls. The system allows organizations to set approval workflows and ensures agents can only access permitted, verified skills.

"Security and governance are key to deploying AI agents in the enterprise," said Pat Lee, vice president of Enterprise Partnerships at NVIDIA. "JFrog's Agent Skills Registry for NVIDIA OpenShell supports security and control for deploying long-running agents."

The companies validated a workflow using NVIDIA cuOpt as the first example of a packaged skill, providing NVIDIA with a governed endpoint for distributing verified AI skills across agent platforms.