Theme Park Pilot — public data mirror

Daily-updated thrill-data.com crowd calendars + (later) historical wait-time aggregates, mirrored by a GitHub Actions workflow in barkeeper/themeparkpilot.

Why a separate repo

The main Theme Park Pilot repo is private. The Flutter app reads these JSONs via raw.githubusercontent.com, which only serves content from PUBLIC repos. Splitting the mirror into its own public repo lets the app fetch without baking any auth token.

Layout

data/
├── index.json                              # which parks have data
└── parks/
    └── <wikiId>/
        └── calendar.json                   # per-park 365-day crowd

<wikiId> is the themeparks.wiki externalId (e.g. MagicKingdomPark).

Schema

data/index.json:

{
  "generatedAt": "2026-05-12T03:00:00Z",
  "scraperVersion": "1.0.0",
  "parks": {
    "MagicKingdomPark": {
      "thrillDataSlug": "magic-kingdom",
      "lastUpdated": "2026-05-12T03:00:00Z",
      "hasCalendar": true,
      "dayCount": 365
    }
  }
}

data/parks/<wikiId>/calendar.json:

{
  "parkExternalId": "MagicKingdomPark",
  "thrillDataSlug": "magic-kingdom",
  "sourceUrl": "https://www.thrill-data.com/trip-planning/crowd-calendar/magic-kingdom",
  "generatedAt": "2026-05-12T03:00:00Z",
  "days": [
    { "date": "2026-05-13", "crowdPercent": 53.3, "predicted": true }
  ]
}

Updates

Updated daily at 03:00 UTC by the thrill-data-scrape.yml workflow in the main repo. PRs to this repo are NOT accepted — the data is machine-generated.

License

Data is derived from public thrill-data.com pages; treat it as "informational only, no warranty". Crowd-level / wait-time numbers are best-effort predictions, not facts.