Connecting to the world

Data is the new gold. The real value of software lies in the data that is behind it. The more data you are able to bring to your users in a way that is easy to understand, the more value your product is bringing. No matter if it is a website about news, webshop, car sales or something else, the data is always a key point. Let's take a car sales page for example. If the users of the page are able to easily find and compare different cars, they will be more likely to use the site and come back but that is not the only data you can collect. You can also collect data about how many cars of each type ar being sold in other shops, how they are priced, how likely they are to break down etc. Not to mention the data about your users, how many people are looking for Scoda on your site, how many are looking for family cars etc...

Having good data is a key point in taking good, and well informed decisions. That is why it is so important to be able to make use of all the different datasources out there. Most of them are closed and protected by companies and many copanies pay a lot of money for access to such data sources. Some use web scraping as a way to get more data but that can both be a fragile solution (break often and easily) and on a gray zone legally.

That is why I have collected a bunch of APIs and MCPs that you can play around with to try to make your own Data driven website, click here and explore the options!

• The different roles of web APIs, MCP servers and agent skills in a development workflow.

• HTTP requests, endpoints, parameters and JSON responses in asynchronous data exchange.

• How data structure, shared identifiers and freshness affect the combination of different sources.

• How authentication, CORS and rate limits affect API access and reliability.

• How MCP permissions and skill instructions shape an agent’s capabilities and behaviour.

• The limits of AI-generated code and explanations, and the role of evidence in technical decisions.

• Read API documentation and assess data sources for a product’s needs.

• Retrieve, transform and combine data from different APIs into a clear interface.

• Diagnose failed requests, handle loading and error states, and keep secret API keys on the server.

• Connect and use an MCP server with appropriate access permissions.

• Inspect, adapt or write agent skills and evaluate their effect on a workflow.

• Verify AI-generated integrations and explain technical decisions using documentation and observed results.

This guide is about combining data from different locations. Use at least two APIs and one MCP server to create a great website. Also explore the skills and make use of at least one of them or create your own. But be careful because skills can also change the behaviour of your LLM so make sure to read it through before you start to use it: