Skip to the content.

Case study: documenting the Open-Meteo Forecast API

Goal: write API documentation that a developer can follow to a first successful request in five minutes, and that stays accurate because I tested every claim before writing it.

Deliverables: an OpenAPI reference with live “Try it out”, a Quickstart, and code samples in curl, Python and JavaScript. This is an independent portfolio sample for a public API, not official Open-Meteo documentation.

How I worked

  1. Explored the API in Postman. I called the forecast endpoint for current conditions, daily summaries and different timezones, and saved each call in a collection.
  2. Tested how it fails. I sent an out-of-range latitude, left out a coordinate and requested an invalid daily variable. I also hit a temporary 503 and recorded it.
  3. Wrote the OpenAPI 3.0 spec and checked it against live responses with Swagger’s “Try it out”, field by field.
  4. Ran every code sample before publishing it, on 11 October 2026.
  5. Published on GitHub Pages, with the spec rendered by Swagger UI.

What testing found

Decisions

What I would improve next

Tools: Postman, OpenAPI 3.0, Swagger Editor and Swagger UI, curl, Python, JavaScript, Git and GitHub Pages.