SYDNEY, Australia. – March 19, 2026 –  JFrog Ltd (Nasdaq: FROG), the Liquid Software company and creators of the JFrog Software Supply Chain Platform, the system of record for software artifacts, binaries, and AI assets, today introduced its JFrog MCP Registry. Expanding on current capabilities in JFrog AI Catalog, the new registry acts as a single source of truth for securely governing Model Context Protocol (MCP) Servers, helping companies transition AI usage from an enterprise-wide compliance and security risk into a competitive advantage.

“Today, developers across the enterprise are rapidly adopting MCP servers from multiple AI tools and vendors, creating a growing challenge for organisations that lack the visibility and control to monitor these connections,” said Yuval Fernbach, CTO, JFrog MLOps. “We’re witnessing a fundamental shift in how software is built and deployed, with AI agents becoming active participants in the software supply chain. This innovation cannot come at the expense of security, visibility, control, or compliance. By establishing a system of record for MCP server usage, and treating them like any other binary asset, organisations can confidently innovate at scale while maintaining the trust and control required across the AIdriven software supply chain.”

The Hidden Risks of Unmanaged MCP Servers

As AI shifts from simple chat interfaces to autonomous, long-running agents, developers rely on MCP servers to act as “enablers of integration,” giving AI models direct access to internal and external enterprise systems, APIs, and data. However, these servers, which act as trusted intermediaries, can also execute arbitrary, potentially malicious code directly on a user’s machine or on remote systems with high privileges. If left unmanaged, they expose organisations to severe risks, including prompt hijacking vulnerabilities, over-privileged access, and credential exposure.

This need for AI governance is backed by Gartner research^[1], stating that security and AI leaders must establish MCPs as the foundational method for agents to communicate with external resources by implementing a centralised MCP server registry, enforcing layered security controls, and defining clear ownership and governance policies.

 Delivering a System of Record for MCP Servers: The JFrog MCP Registry

The new JFrog MCP Registry provides a system of record and AI infrastructure trust layer for all MCP Servers, agent skills, models, and agentic binary assets. By treating MCP servers with the same rigorous security standards as software packages, the JFrog MCP Registry helps eliminate blind spots across the AI software supply chain. At its core the JFrog MCP Registry is designed to bring:

• Native security by design to proactively block the download and execution of malicious or non-compliant MCP servers, otherwise pulled naively by humans or AI agents, rather than waiting for an issue to occur and remediating it after the fact. 
• Centralised governance and management enabling developers to instantly access a registry of pre-approved local and remote MCP servers directly from their Integrated Development Environments (e.g., Claude Code, Cursor, VS-Code). 
• Enterprise-grade policy enforcement on every agentic workflow, replacing “blind trust” with granular control, by treating every MCP server as a governed artifact with centralised discovery, configuration and project-level permissions management alongside all other AI models and software artifacts in a unified AI Catalog. 
• Platform universality, which allows companies to seamlessly manage agent ecosystems from private marketplaces and across vendors, enabling teams to seamlessly switch coding agents without ever needing to rebuild their secure system of record.

The JFrog MCP Registry is available immediately as part of JFrog AI Catalog. To learn more about how it works, readthis blog, visit https://jfrog.com/ai-catalog/mcp-registry, or register for the “The Right Tools for the Job: Securing Your AI Agents” webinar on March 31 at 10 AM PST.