Search for a command to run...
Series
In this series, I will be explaining AI in a structured, step-by-step manner, helping readers "decode" complex topics.
Gateway, API gateway, AI-native gateway: the concept ladder, the real players, and where AgentGateway actually stands next to AWS, Azure, Kong, and LiteLLM.
The biggest LLM-consumption release yet: a purpose-built UI, per-token cost attribution, and virtual models that reroute traffic without touching a single client.
One file runs everything — binds, listeners, routes, backends, policies. Here's how to read it top to bottom, and why the same skeleton powers an MCP gateway, an LLM gateway, and a plain proxy.
No Rust, no Docker, no compiling. Grab the Windows binary, write a tiny config, and watch a real MCP tool server show up in your browser.
One open-source proxy for LLMs, MCP tools, agents, and your regular APIs, plus the data-plane and control-plane vocabulary you'll keep running into.
Introduction In my earlier blogpost, we built a Coordinator/Dispatcher Agent that routes tasks to specialized workers. We also read about how agent patterns mirror microservice patterns—API Gateways, Sagas, and Domain-Driven Design. But there's one c...