data: non-SAT place categories are extremely thin across the dataset's 378 cities

[yakew7]

Problem

The dataset now spans 378 distinct cities, but that coverage is almost entirely sat_centre.json (609 records). Every other category is thin relative to that footprint:

File	Records
library.json	27
gov_offices.json	14
airport.json	3
foreign_lang_exam_centre.json	2
other_places.json	1

That's 47 non-SAT records total spread across a 378-city dataset - most cities with an SAT centre have zero libraries, government offices, or anything else pinned for the students already using that page.

Task

Add libraries, government/civic offices (passport offices, RTOs, embassies, post offices), airports, and foreign-language exam centres (IELTS, TOEFL, Goethe, DELF, etc.) for cities that already have SAT centre coverage - that's where students are already looking. Prioritize whichever cities you have direct knowledge of, or the highest-population cities already in the dataset.

Data requirements

Follow data/CONTRIBUTING.md exactly - correct file per type, unique id following the <city-prefix>-<type>-<number> convention, real coordinates (from the place's own Google Maps listing, not a city-center placeholder - see #227 for why that matters), and a valid gmaps_link.

How to contribute

See CONTRIBUTING.md and data/CONTRIBUTING.md. npm run validate must pass before opening a PR.

Acceptance criteria

• At least one of the thin categories gets meaningfully more coverage across cities that already have an SAT centre entry