The Model Context Protocol is how AI assistants connect to calendars, files and databases. What it is, why it caught on, and what it lets a small business do.
An AI assistant on its own only knows what you paste into it. To be useful in a business, it needs to see your bookings, your files or your orders, and sometimes take an action. The Model Context Protocol, MCP, is the most common way to make that connection.
The simple version
Think of MCP like a standard plug. Before it, every AI tool needed its own custom connection to every app. With MCP, a business system offers one “server” that describes what it can do, and any AI tool that speaks MCP can use it.
An MCP server might offer actions like “list tomorrow’s bookings”, “look up a client by email” or “create a draft invoice”. The AI decides when to use them; the server decides what is allowed.
Why it caught on
Anthropic released MCP as an open standard in late 2024. Other AI companies adopted it, and in late 2025 Anthropic donated it to the Agentic AI Foundation, a fund under the Linux Foundation co-founded with Block and OpenAI. A connection built once now works across many AI tools rather than being tied to one.
On July 28, 2026, MCP released its largest update yet, which made servers simpler to host and tightened how tools sign in.
What it lets a small business do
- Ask your assistant “who hasn’t confirmed for Saturday?” and get an answer from your real booking data.
- Have AI draft a reply to a client using their actual file, not a guess.
- Let a coordinator say “move the U12 practice to field 2 and tell the coach” and review the change before it goes out.
- Pull this week’s sales and stock into a short summary every Monday.
What to get right
- Least access. A server should offer only the actions the job needs. Reading bookings and cancelling them are different permissions.
- Approval for actions. Reading can be automatic; changing, sending or deleting should ask a person first.
- Logs. Every action should be recorded.
- Trusted servers only. Installing a random MCP server is like installing a random browser extension with access to your data.
MCP is the plug. What matters is deciding carefully what is on the other end of it.
Sources
From the DYOL blog. Photo by Lucian Alexe on Unsplash.
Keep reading
All posts
MCP’s biggest update yet: what the July 2026 spec changes
The Model Context Protocol dropped sessions, tightened sign-in rules and made long tasks official. What changed, and why it makes AI connections easier to run.

What’s new in Claude Code: the late-summer releases that matter
Sessions that talk to each other, design in the terminal, background computer use and a way to test plugins. Here is what changed and why it matters.

What’s new in AI this autumn, in plain language
Cheaper fast models, agents that run on a schedule and new rules on labelling AI. A short tour of what changed and what it means for a small team.
A free 20-minute conversation, then an honest assessment. You’ll know what it costs and how long it takes, in writing, at a fixed price.
Or email lloyd.delacruz@outlook.com. We reply within 24 hours.
