Follow ZDNET: Add us as a preferred source<!–> on Google.
ZDNET’s key takeaways
- Solo’s Agentregistry lets you track AI agents.
- Solo offers open-source AI agent tools to help users and programmers.
- The platform could become the foundation for managing AI agents.
AI agents have become a big deal in a very short time. That’s great, but how can anyone possibly keep track of them all? At Kubecon North America 2025 in Atlanta, the open-source, cloud-native company Solo.io presented an answer. Idit Levine, founder and CEO of Agentregistry Solo, donated Agentregistry to the Cloud Native Computing Foundation (CNCF) during her keynote speech.
Also: Cloud-native computing is poised to explode, thanks to AI inference work
This gesture followed Solo’s earlier donation of AgentGateway, the first data plane built for AI agents, to the Linux Foundation at the Open Source Summit Europe 2025 in August. AgentGateway is an open-source project built to connect, secure, and observe agent-to-agent and agent-to-tool communication.
Complete agentic infrastructure stack
Agentregistry is an open-source platform that delivers a centralized, trusted, and curated registry for AI applications, agents, and Agent Skills. The goal is to streamline the discovery, validation, and operationalization of AI agent skills across diverse frameworks and platforms. In short, the platform is meant to be a “single source of truth” for agent developers and users.
Agent Skills, meanwhile, is a framework from AI giant Anthropic. The framework consists of folders containing scripts and resources that you can load into an agent as needed. Think of these resources as instructions and recipes for AI agents, and you’ve got the idea.
Also: Kubernetes, cloud-native computing’s engine, is getting turbocharged for AI
Agentregistry is part of a broader push by Solo.io to provide a complete agentic infrastructure stack, alongside projects such as Kagent and AgentGateway. Kagent is an open-source framework for running AI agents in Kubernetes.
Levine, a former professional basketball player, emphasized that “we’re at an inflection point in AI where agents are moving beyond simple tool calling to dynamically writing and executing their own code.” She noted that the future will see agents not only connecting to tools but also autonomously generating, executing, and sharing new skills with other agents, creating workflows previously impossible in traditional environments.
Essential for enterprise adoption
This platform goes beyond the community-created Model Context Protocol registry. It provides layers of security, governance, and rich metadata management, which are considered essential for enterprise adoption. According to Solo.io’s chief product officer, Keith Babo, Agentregistry “provides the complete agentic infrastructure stack that enterprises need to bring AI to production in Kubernetes.”
With context-aware features that integrate with other Solo.io offerings, Agentregistry is positioned as vital infrastructure for running AI workflows in cloud-native environments. Solo.io’s approach is to make the Kubernetes ecosystem “ready for the future” of agentic AI, enabling easier management and secure distribution of modular AI capabilities.
Also: Why even a US tech giant is launching ‘sovereign support’ for Europe now
This is a big gap that Solo is trying to fill for organizations that need to securely publish, discover, share, and version AI agents and skills within a trusted environment. There is a real need here. Thanks to their years of experience, Solo already has a good reputation in the cloud-native space. Based on that background, and the word of mouth I heard at Kubecon, I expect Agentregistry to gain traction quickly.
Wild West
I think we’re still in the Wild West phase of deploying and integrating AI agents. Solo hopes to be the sheriff who brings law and order to this field.
