Brian Best
Brian Best
← Back to blog

MCP Is the Real Unlock for Agentic Engineering

mcpagentic-aitooling

A year ago, every "agent" I built was a bespoke pile of tool definitions, retry loops, and JSON-schema babysitting. Each new integration meant teaching the model a new dialect, hand-rolling auth, and praying the function-calling format didn't drift between providers. It worked, but it didn't compose.

Then MCP showed up and quietly turned the problem inside out.

The shift

The Model Context Protocol gives tools a standard shape — a server speaks MCP, a client speaks MCP, and the model on the other end doesn't care who wrote either side. That sounds boring. It is not boring. It is the difference between "I built an agent that can do X" and "I plugged my agent into the ecosystem of things that already do X."

The first time I swapped a custom Confluence integration for an off-the-shelf MCP server and watched it Just Work, I felt the same thing I felt the first time I used npm install instead of vendoring a library by hand.

What changed for my team

  • Tool development decoupled from agent development. A junior dev can ship an MCP server in an afternoon. They don't need to know anything about prompt engineering, model selection, or context windows.
  • Tools became testable in isolation. MCP servers are just servers. You unit-test them like servers. You don't need an LLM in the loop to know the tool works.
  • Composition got cheap. We routinely run agents with eight or ten MCP servers attached. Six months ago that would have been a custom orchestration nightmare; today it's a config file.

What I still don't love

MCP's discovery story is weak. There's no good answer yet for "which of my 47 connected tools should the model reach for in this context?" Most teams I talk to end up writing a meta-layer that filters tool exposure per-task. That's a band-aid, not a fix.

Permission scoping is also early. The protocol punts most of the auth model to the implementer, which means every team is reinventing the same approval flows. I expect this to converge in the next year.

The bet

If you're starting an agent project in 2026 and you're not building on MCP, you're choosing to do extra work for no benefit. The protocol won. Build on it.