Cloud native application networking company Solo.io today announced Agent Gateway, an open source data plane optimized for agentic AI connectivity in any environment. Agent Gateway provides drop-in security, observability, and governance for agent-to-agent and agent-to-tool communication and supports leading interoperable protocols, including Agent2Agent (A2A) and Model Context Protocol (MCP).

When developing and deploying AI agents, organizations face the challenge of supporting multiple rapidly evolving protocols across fragmented teams, environments, and agent development frameworks. Agent Gateway provides a unified data plane for agent connectivity, supporting A2A and MCP, with the ability to automatically integrate an organization’s existing REST APIs as agent-native tools. A built-in developer portal provides tool providers and agent developers with a single pane of glass to discover, configure, and monitor agent-to-agent and agent-to-tool connectivity. The Agent Gateway data plane seamlessly integrates with popular agent frameworks, including LangGraph, AutoGen, Agents SDK, kagent, and Claude Desktop. It runs wherever agents run, including bare metal, virtual machines (VMs), containers, and Kubernetes.

As agent development practices mature, the industry is finding that smaller, focused agents, aligned with specific goals or tasks, perform better than a single, monolithic agent. Just like microservices created the need for a service mesh to address cross-cutting concerns at the connectivity layer, agents require an Agent Mesh to solve common security, observability, tenancy, and guardrail concerns. The release of Agent Gateway builds on the robust open source foundation of kgateway and Ambient Mesh to create an Agent Mesh architecture for AI use cases spanning LLM consumption, inferencing, tool calling, and agent-to-agent communication. Agent Mesh enables seamless security, observability, discovery, and governance across all agent interactions, regardless of how the agents are built or where they are deployed.

“Agentic AI is transforming how organizations build and deliver applications, but long-term success requires infrastructure that transcends today’s rapidly changing landscape,” said Idit Levine, founder and CEO of Solo.io. “Using industry standard protocols like A2A and MCP helps organizations future-proof their AI applications by ensuring interoperability with any LLM or agent framework. Agent Mesh brings these standards together with the leading open source gateway and mesh to form the only comprehensive AI connectivity stack for agentic applications.”

Agent Mesh seamlessly integrates Agent Gateway into the AI connectivity plane to support any MCP tool server, agent framework, LLM, and runtime environment used in an organization’s agentic architecture, providing:

  • Comprehensive, secure-by-default architecture with agent identity and mTLS.

  • Multitenant access boundaries and controls for agents and tools across teams and environments.

  • Standard agent connectivity with A2A and MCP, with the ability to automatically integrate existing REST APIs as MCP-native tool servers.

  • Automated collection and centralized reporting of agent telemetry, including metrics, tracing, and logging.

  • A self-service agent developer portal supporting discovery, configuration, observability, and debugging tools for agents and tools.