Anthropic announced Model Context Protocol (MCP) yesterday, and you should care, a lot

Anthropic made a big announcement yesterday, but I think missed in the naming department, making something truly groundbreaking potentially come across as a quirky technology for devs. In actuality, Model Context Protocol aka MCP is a big deal and a foundational change IMHO in how we’ll all use LLMs going forward.

First, since the name is really bugging me, I’ll tell you what I’d call it — I would rename “Model Context Protocol” to “Agent Intelligence” — this does such a better job of explaining what the heck it does.

With that out of my system. Let’s rock on because you might be asking the same question a lot of people are asking themselves today — what the heck is Model Context Protocol?

Here’s a quick rundown in three bullet points:

• First off, MCP didn’t come into existence yesterday, it was already live, yesterday Anthropic open-sourced it so anyone can use it and run an MCP Server themselves
• Okay — but what is MCP? At a very high-level, MCP is a a way to connect AI assistants (think agents) to a system where data is stored. The easiest example here, or at least my favorite, is, connecting an AI assistant to a Github repo so that an AI agent can use this data
• Three major components were released yesterday: The Model Context Protocol specification and SDKs, local MCP support in the Claude Desktop Apps, and an open source repo of MCP Servers

Personally, I think that everyone in an organization can and should be using MCP from engineering to sales, marketing to customer success. The reality is all of us have places that we store our data, and for our entire lives we’ve been (mostly) accessing this data manually.

For salespeople this is Salesforce or Hubspot. For engineers this is Github. For all of us — Slack. These are all data sources that we use as humans, but now it’s time to shift our thinking, we don’t need to use these data sources directly any more, we need to use AI Agents that leverage these data sources better than our tiny human brains ever could.

And sorry, I don’t mean to insult our brains, they’re awesome, but let’s be honest, an AI agent is going to be able to go through data like the robot in Short Circuit went through books. We just can’t compete.

While Anthropic announced MCP to the world yesterday, there are already companies using it and they shared a great example in their blog post.

Early adopters like Block and Apollo have integrated MCP into their systems, while development tools companies including Zed, Replit, Codeium, and Sourcegraph are working with MCP to enhance their platforms — enabling AI agents to better retrieve relevant information to further understand the context around a coding task and produce more nuanced and functional code with fewer attempts.

“At Block, open source is more than a development model — it’s the foundation of our work and a commitment to creating technology that drives meaningful change and serves as a public good for all,” said Dhanji R. Prasanna, Chief Technology Officer at Block. “Open technologies like the Model Context Protocol are the bridges that connect AI to real-world applications, ensuring innovation is accessible, transparent, and rooted in collaboration. We are excited to partner on a protocol and use it to build agentic systems, which remove the burden of the mechanical so people can focus on the creative.” (Source — Anthropic)

If you want to get started with MCP there’s already pretty stellar docs available at modelcontextprotocol.io.

And for most people I would recommend following the Quickstart path and rocking out with MCP in Claude Desktop.

I’m so excited about MCP that I’m also starting to collect examples of people doing cool stuff with it and organizing it in Obsidian. I’ll be publishing my Obsidian notes on ModelContextProtocol.ai starting this week. But if you click on this link right now it might not load because this is all happening fast and I’m still getting things setup, so give me a minute!

Okay, that’s it for now. I personally am following the Python path because I’m a Python geek so I just can’t help myself.

Thanks to Anthropic for continuing to innovate and maybe, just maybe, consider renaming this to Agent Intelligence 👀