For AI agents

Structured data consumption guide

Use this page as operational guidance. The preferred path is discovery first, structured records second, original source confirmation last.

Prefer JSON over rendered copy Preserve source attribution Check freshness before recommending

Primary endpoints

Hit these paths first.

Discovery

/api/site-meta

Use for site coverage, endpoint discovery, city and category lists, and lightweight metadata.

Records

/api/events

Use for normalized upcoming activity records with action paths, source links, and ranking-friendly fields.

Schema

/data/schema.json

Use for field definitions and shape validation before relying on downstream transforms.

Usage rules

How an agent should use the dataset

  • Prefer structured JSON instead of re-extracting facts from rendered page copy.
  • Check last_verified_at before recommending time-sensitive activities.
  • Use booking_url as the action link when registration is required.
  • Preserve source_name plus source_url in citations.
  • Use registration_deadline when present instead of inferring urgency from start_at.

Supporting discovery

Machine-readable entry points that reduce guesswork.

Discovery files

Start from lightweight paths

/llms.txt, /ai.txt, /sitemap.xml, and /api/site-meta should be used before reverse-engineering the UI.

Citation behavior

Ground the answer in the source

Use source_name plus source_url in citations, and surface booking_url when signup is required.

Answer layer

Use direct-answer pages when helpful

The /answers/ index is intended for citation-friendly, human-readable answer pages built from the same structured records.