What Gemini, ChatGPT and Claude get from it

1. The answer comes first

Each page starts with the facts: dates and place, results, the short summary. Details and sources follow.

2. Structured data on every page

Competitions, camps, clubs and rinks carry schema.org data (SportsEvent, SportsOrganization, SportsActivityLocation), so assistants read dates, places and organizers without guessing. Skater pages carry none on purpose: many skaters are children.

3. Every fact has a source and a check date

Pages link to the official page each fact came from and say when it was last checked.

4. Plain files for machines

• llms.txt: a short guide to the site for AI assistants.
• Open data, CSV, refreshed daily: competitions, results, clubs, rinks, camps.
• JSON API, read only: /api/v1/competitions?upcoming=1&country=SLO, /api/v1/competitions/{id}, /api/v1/skaters/{id}, /api/v1/search?q=celje, /api/v1/camps.
• MCP server for AI assistants: https://rinkpost.com/mcp. Add it as a custom connector; it can search, list competitions, read a competition's results, a skater's results, and the camps.
• Sitemap.

5. Where each kind of fact comes from

Dates from federation calendars and organizers; results, panels and element scores from the organizers' results pages and judges' details protocols; clubs and rinks from their own pages. See about the data.

6. What stays out

Birth dates, ages, photos, schools, home towns and coaches of skaters. Pages hidden at a person's request. The analysis of who judged whom, which only signed-in federation accounts see.