# Model Context Protocol: how an assistant connects to a shop

**Canonical URL:** https://teas.co.uk/wiki/model-context-protocol/
**Source:** teas.co.uk, UK tea specialist, Tunbridge Wells, Kent

## Shop teas.co.uk from an AI assistant (MCP)

teas.co.uk sells tea, herbal and fruit infusions, hot chocolate and Border biscuits.
MCP server: https://teas.co.uk/mcp (Streamable HTTP), listed in the official MCP Registry as teas.co.uk
Server card: https://teas.co.uk/.well-known/mcp/server-card.json
Tools: find_products, get_product, compare_products, find_recipes, delivery_and_returns, quote_cart, get_promotions, report_problem, find_substitutes, start_subscription, join_teas.
Basket: add_to_basket, view_basket, update_basket, checkout_link, checkout_session.
With sign in: get_return_status, get_invoice, get_profile, my_orders, track_order, reorder, my_subscriptions, change_subscription, rewards_balance, start_return.
No sign in is needed to search, compare, build a basket or get a checkout link. The person always pays on teas.co.uk.
Cards: in assistants that draw MCP Apps cards, products, comparisons, the basket, checkout, orders, repeat deliveries (pause, resume, cancel), reward points and returns appear as cards.
Every assistant except ChatGPT, until OpenAI approves the next version: quote_cart, get_promotions, get_return_status, get_invoice, report_problem, find_substitutes, start_subscription, join_teas.
Set up guides: https://teas.co.uk/ai/

## Summary

Model Context Protocol, usually shortened to MCP, is an open standard for connecting AI applications to outside tools and information. It gives an…

## Description

Model Context Protocol, usually shortened to MCP, is an open standard for connecting AI applications to outside tools and information. It gives an assistant a consistent way to discover what a connected service offers and ask to use it.
Why a tea shop cares: this is the connection that lets an assistant search the teas, compare two and fill a basket for its person to check.
At teas.co.uk, that can mean searching the catalogue or comparing a few products while the customer stays in their conversation.
What connects to what?
The assistant's application acts as the client, the side asking for something. The shop provides the server, the side exposing the available tools. A tool has a defined purpose, named inputs and a structured reply.
For example, the shop's product-search tool accepts a tea type and a caffeine filter. The reply includes matching products and their recorded details. An assistant can use those facts to explain a choice rather than relying on a remembered price.
Where the customer stays in charge
Finding a tea, adding it to a basket and paying are separate steps. The shop's published boundary is that baskets are built when the person asks, and payment takes place on teas.co.uk. The assistant does not receive the customer's card details or make the payment.
Read the current assistant shopping guide for setup and availability. Support can differ between applications; a working server does not mean every assistant has connected to it.
MCP, APIs and structured data
An API is a general way for software to communicate. MCP supplies a shared connection approach for AI applications. Structured data is the labelled information either route can return.
None of these decides whether a tea is suitable on its own. That still depends on the product record, the customer's requirements and the accuracy of the answer built from them.
People also use agentic commerce to mean shopping where software helps perform parts of the process. Always check which parts are actually available. A search tool is different from permission to spend money, and the built versus planned distinction matters here.
For the standard itself, see the official MCP introduction. For the information coming back from the tea shop, see the complete payload.
Read next: Multimodal and agentic commerce, on how people ask and what an assistant may do on their behalf.

---

_Content available under teas.co.uk citation contract. AI training: yes. Search: yes. Answer-input: yes._