At NVIDIA GTC, JFrog Ltd. unveiled its new Agent Skills Registry, marking a significant step toward securing AI-driven enterprise environments. The company introduced this solution as part of its broader platform strategy to enable trusted, scalable, and governed AI agent operations. Notably, the new registry has been validated through initial integration with NVIDIA, further strengthening its enterprise readiness.
To begin with, the JFrog Agent Skills Registry is designed to provide governance and verifiable trust for AI agents operating across modern software supply chains. As organizations increasingly deploy autonomous agents, the need for a secure and policy-driven infrastructure becomes critical. Therefore, JFrog aims to address this gap by offering a centralized system that ensures secure execution and management of AI-driven workflows.
In addition, the registry supports the NVIDIA Agent Toolkit, including NVIDIA NemoClaw, an open-source execution environment built for developing and deploying secure, long-term autonomous agents. At the same time, JFrog Artifactory will function as a unified registry for AI models and agent skills within the NVIDIA AI-Q Blueprint ecosystem. As a result, enterprises can manage both software artifacts and AI assets within a single, governed platform.
“AI agents are essentially redefining the way software is built and executed, but without a dedicated trust layer to enforce governance and secure workflows, they introduce significant business risk,” said Gal Marder, Chief Strategy Officer at JFrog. “Just as a malicious software package can compromise an application, unverified expertise can lead an agent to perform malicious actions. To deploy autonomous agents at scale, organizations must move beyond blind trust. Working closely with NVIDIA’s Enterprise AI Factory team, we are defining a trusted system of record to store, scan, and govern all agent binary assets across the entire software supply chain.”
Meanwhile, the rapid rise of AI technologies has accelerated the adoption of skill-based autonomous agents. However, without a standardized infrastructure layer, organizations face serious risks related to security, compliance, and privacy. For instance, recent OpenClaw manipulations and breaches have highlighted the dangers of ungoverned AI systems. Consequently, JFrog’s solution focuses on enforcing strict policies and controls to mitigate these threats.
Furthermore, the platform introduces several capabilities to strengthen enterprise AI operations. It enables organizations to improve security and governance by scanning and blocking malicious or vulnerable components using a single source of truth. Additionally, it supports the safe scalability of long-term autonomous agents without compromising compliance requirements. At the same time, it empowers developer innovation by enabling secure, continuous agentic workflows across the enterprise.
“Security and governance are essential for deploying AI agents across the enterprise,” said Pat Lee, Vice President, Enterprise Partnerships at NVIDIA. “The JFrog Agent Skills Registry for NVIDIA NemoClaw supports security and control for deploying agents long-term to help boost enterprise productivity with powerful new AI tools.”
Moreover, the integration between JFrog and NVIDIA establishes a secure, unified registry for managing AI assets such as agent skills, models, MCP servers, and software packages. Through this collaboration, both companies validated a workflow that uses JFrog Artifactory as a centralized skills registry. For example, NVIDIA cuOpt has been introduced as one of the first predefined skills within this ecosystem. As a result, enterprises gain access to a governed deployment model that enforces increasing security layers from team-level usage to enterprise-wide adoption.
In terms of key features, the new offering includes validation for the NVIDIA AI-Q Blueprint, ensuring robust lifecycle management and governance for AI agents. Additionally, JFrog Artifactory integrates natively with the NVIDIA NemoClaw runtime, enabling secure handling of private and scanned assets. The platform also acts as a centralized system of record through the JFrog AI Catalog, allowing organizations to track and verify the provenance of all AI-related components.
Finally, the solution emphasizes proactive security by automatically scanning, verifying, and signing AI skills upon upload. This ensures that vulnerabilities, malicious payloads, and compliance risks are identified before deployment. Combined with policy-based governance and isolated runtime environments, JFrog enables enterprises to safely scale AI adoption while maintaining strict control over their software supply chain.
Overall, JFrog’s latest innovation highlights the growing importance of trust, governance, and security in the rapidly evolving world of AI-powered enterprise operations.
Recommended Cyber Technology News:
- Emovid Launches Verified Human Communication Platform
- BitSight Appoints John Clancy as Chief Executive Officer
- Nutanix NetApp Alliance Drives Modern Cloud Platform Growth
To participate in our interviews, please write to our CyberTech Media Room at info@intentamplify.com
