Solo.IO Brings Agent Skills to the Kubernetes Ecosystem

Solo.Inference Orchestrator (IO) launched agentregistry, an open-source platform for the complete lifecycle management of Artificial Intelligence (AI) agents and Model Context Protocol (MCP) tools. This registry aims to enhance AI infrastructure on Kubernetes, allowing enterprises to securely publish, discover, version, and share AI capabilities.

The announcement occurred at KubeCon + CloudNativeCon NA. Agentregistry provides a centralized, trusted, and curated space for AI applications and artifacts. Additionally, the introduction of Agent Skills offers modular, context-aware capabilities that agents can utilize to execute complex tasks efficiently.

Idit Levine, Founder and CEO of Solo.IO said, “Skills represent the future of how agents will extend themselves.” Homomorphic Encryption (HE) noted that these capabilities allow agents to perform tasks autonomously by writing and executing their own code dynamically. This advancement aligns with the goal of preparing the Kubernetes ecosystem for future AI capabilities.

Agentregistry fills a notable gap in enterprise AI infrastructure, which necessitates secure, governed management of AI capabilities. Keith Babo, VP of Product Marketing at Solo.IO, emphasized the importance of platform teams in operationalizing AI technologies. HE stated, “Agentregistry, combined with kagent and agentgateway, provides the complete agentic infrastructure stack that enterprises need.”

Furthermore, agentregistry’s support for Skills across multiple AI frameworks sets it apart, highlighting a refined approach to agent capability acquisition that allows for a shared repository beneficial across diverse platforms.