MCP is the boring center that makes any of this work. The cliché is that MCP is “USB-C for AI”: one open protocol, any software. Like most analogies, it’s about eighty % proper. The half that issues is the eighty: I wouldn’t have to put in writing a customized adapter for each system the agent talks to. One MCP server per software and each agent talks to all of them the identical manner.
Typed handoffs between brokers are my very own structure, layered on high of MCP quite than offered by it. Every agent writes a typed artifact the subsequent agent reads. Every handoff is logged with provenance. When one thing went unsuitable six phases in, I might replay the chain. With out that self-discipline, a multi-agent pipeline is a debugger’s worst day. the check plan is unsuitable. You can not inform whether or not the error got here from the Figma learn, the necessities interpretation or the ticket scaffolding. With it, I might level at precisely which stage went sideways and which inputs it was taking a look at when it did. The sample lives in a public MIT-licensed reference implementation for any reader who desires to run it.
The sixteen-minute quantity is the advertising and marketing quantity. I ran the complete chain finish to finish in about sixteen minutes on an artificial net-new display, Figma in, automation suite out. That repeated throughout my runs; it’s not a demo trick. However sixteen minutes is the a part of the story most enjoyable to inform and least helpful to be taught from. It’s what will get quoted within the all-hands. The hours that come after, when a human critiques every handoff, are the place the work truly lives.