DocsNorthwind walkthrough

OData to MCP walkthrough

Convert Northwind OData into MCP tools, prompts, resources, and UI.

This guide shows the full DataTether setup path with a public read-only OData service. Use it to practice before connecting SAP Gateway, SAP S/4HANA, or another enterprise OData endpoint.

Start the walkthroughRequest guided accessView OData sample services

Demo endpoint

https://services.odata.org/V2/Northwind/Northwind.svc/
No SAP credentials required
Read-only demo service
Good for MCP tool and UI training
Same flow as SAP OData setup

Step-by-step procedure

Each form, field, and decision in order.

Follow this in order for a clean demo: connect the sample endpoint, keep tools read-only, add helpful context, then test the MCP client.

Step 1

Open Add Connection

Dashboard -> Add Connection

Start a new DataTether connection. This opens the same flow you will use later for SAP Gateway or S/4HANA OData services.

Use these fields

  • Connection Name: Northwind Demo
  • OData Service URL: https://services.odata.org/V2/Northwind/Northwind.svc/
  • Auth Type: No Auth
Step 2

Discover the service metadata

Discovery

DataTether reads the OData metadata document and turns entity sets, fields, keys, and functions into a configurable capability inventory.

Use these fields

  • Confirm entities such as Customers, Orders, Products, Categories, Suppliers
  • Open View Metadata if you want to inspect fields before selecting tools
Step 3

Select MCP tools

Configure -> Tools

Choose the entities and operations the AI client can call. For Northwind, start with read-only tools so the demo stays safe and easy to explain.

Use these fields

  • Enable Customers, Orders, Products, Categories
  • Enable Filter/List, Get, Count, and Search where available
  • Disable Create, Update, and Delete for this read-only demo
Step 4

Enable MCP resources

Resources

Resources provide schema and documentation context without executing business operations. They help the AI understand fields before it calls a tool.

Use these fields

  • Keep auto-generated metadata resources enabled for selected entities
  • Add a custom resource for demo rules, field meanings, or filter examples
  • Save resource changes
Step 5

Create MCP prompts

Prompts

Prompts make repeatable workflows easier for users. Add arguments so the same template works for different customers, date ranges, or product categories.

Use these fields

  • Prompt: Find customer orders for {{customerId}}
  • Prompt: Summarize product inventory for {{categoryName}}
  • Use Auto Create to define placeholder arguments
Step 6

Map UI views

UI

Build a visual response surface by binding OData fields into tables, cards, KPI widgets, and dashboards.

Use these fields

  • Create a customer orders dashboard
  • Alias data sources as customers, orders, and products
  • Map CustomerName, OrderID, OrderDate, ProductName, Quantity, and UnitPrice
Step 7

Connect an AI client

Connect

Copy the MCP endpoint into the client you want to test. Start with read-only prompts, then inspect tools, resources, and UI views from the AI client.

Use these fields

  • Choose ChatGPT, Claude, MCP Inspector, Manus, or another MCP client
  • Select Streamable HTTP where supported
  • Test with: List available tools and describe what each tool can do

Screenshot order for the docs page.

Capture these wizard screens when you have a live local or staging session. Resources, Prompts, and UI appear after Save & Continue from Configure, so capture them as separate steps.

Step 1 Connection: Northwind URL and No Auth
Step 2 Discovery: metadata success with entity and function counts
Step 3 Configure: Customers, Orders, Products, Categories tools enabled
Step 4 Resources: auto-generated metadata resources enabled
Step 5 Prompts: reusable prompt with {{customerId}} argument
Step 6 UI: dashboard or card mapping fields to widgets
Connect page: MCP URL and client setup checklist
AI client test: tools/list result or first safe read prompt

FAQ: MCP, OData, tools, prompts, and resources

What is OData to MCP?

OData to MCP is the process of turning an OData service contract into Model Context Protocol tools, resources, prompts, and UI surfaces that AI clients can safely discover and use.

What MCP tools does DataTether generate from OData?

DataTether can generate list/filter, get, count, search, create, update, delete, function/action, and service-info tools depending on the OData metadata and admin-selected operations.

Why use MCP resources with OData tools?

MCP resources expose readable context such as service metadata, entity schemas, field descriptions, and custom guidance. They help AI clients understand the data model before calling tools.

How are MCP prompts different from MCP tools?

A prompt is reusable instruction text with arguments. A tool is an action the AI can call. Prompts guide the user and model; tools perform the approved OData operation.

Can the same setup be used for SAP OData?

Yes. Northwind is a public read-only demo service, but the same DataTether flow applies to SAP Gateway, SAP S/4HANA OData, Dynamics OData, Salesforce OData, and custom enterprise OData services.

Try it with your own service

You just turned Northwind into MCP tools. Now do it with your real OData service.

Same steps. Your real data. Live today.

The same steps work for SAP, Odoo, Oracle, Dynamics, or any OData v4 endpoint. Connect once, review the generated tools, and go live with governed AI actions the same day.

Months of dev time eliminated

Auto-generated tools replace weeks of custom MCP server development.

No infrastructure to run

Fully managed. Nothing to install, host, or maintain.

Instant dashboards and UI

Live tables, KPIs, and review screens — no frontend developer needed.

One setup, every AI client

ChatGPT, Claude, Manus, custom agents — all from the same connection.

A good first step is small: one service, one user outcome, one visible control boundary.

Request guided access