How this works

Oman publishes real statistics — but no developer layer: no developer-facing API, no versioned bulk downloads, no format guarantees. This project is that layer, built as a companion to the official portals, not a replacement.

  1. Official sources only. Every dataset is fetched from an officially published file (NCSI, ministries, regulators). Nothing is scraped that wasn't published deliberately.
  2. Every snapshot archived. The exact source file of every fetch is committed under raw/ — every number traces back: JSON → git commit → archived snapshot. Most snapshots are an official source file; where a value had to be reconstructed rather than published — part of the fuel-price history, for which no official machine-readable archive exists — the row's own source column says so.
  3. Validation gates publication. Schema, ranges, and row-count sanity are checked before anything is published. A broken source never breaks published data — the last good version stays.
  4. Honest freshness. A dataset past its update window is flagged "stale": true — never silently outdated.
  5. Bilingual by contract. Every dataset carries Arabic and English titles; Arabic is stored unescaped. A dataset without both doesn't ship.

Use it

curl https://omandata.dev/v1/datasets.json

Static JSON, CORS open, no keys, no rate limits beyond the CDN's. Bulk data as CSV/Parquet lives in the pipeline repo.

Licensing & attribution

Code is MIT. Each dataset keeps the license of its official source (Open Government License – Oman for government data). This is an independent project by Nasser Albusaidi, not affiliated with any government entity.