`animedex ghibli`

The Ghibli backend is a bundled, offline snapshot of the public Studio Ghibli API. High-level commands read animedex/data/ghibli.json directly, so they are deterministic, do not need network access, and do not consume upstream capacity. The raw animedex api ghibli passthrough remains available when a caller explicitly wants the live mirror.

References

Live API	https://ghibliapi.vercel.app/
Source repository	https://github.com/janismdhanbad/studio-ghibli-api
Python module	`animedex.backends.ghibli`
Rich models	`animedex.backends.ghibli.models`

Backend: Studio Ghibli API snapshot bundled with animedex.
Rate limit: not applicable for high-level commands; reads are local file reads.
Auth: never. The live mirror and bundled snapshot are anonymous read-only data.
Snapshot: captured on 2026-05-09 UTC from https://ghibliapi.vercel.app/{films,people,locations,species,vehicles}.

Films — `films()`

films lists the 22 bundled film records in snapshot order and supports local filters:

animedex ghibli films --director "Hayao Miyazaki" --min-rt-score 95 --jq 'map(.title)'
# => ["Castle in the Sky", "Kiki's Delivery Service", "Spirited Away"]

animedex ghibli film 2baf70d1-42bb-4437-b551-e5fed5a87abe --jq '{title, director, rt_score}'
# => {"title": "Castle in the Sky", "director": "Hayao Miyazaki", "rt_score": "95"}

Film records project to the cross-source Anime shape via animedex.backends.ghibli.models.GhibliFilm.to_common(), carrying the Ghibli UUID in ids["ghibli"] and the Rotten Tomatoes score as a 0..100 rating.

People — `people()`

people lists character/person records and accepts a positional optional name filter plus --gender, --film-id, and --species-id:

animedex ghibli people Haku --jq '.[0] | {name, age, eye_color, source: .source_tag.backend}'
# => {"name": "Haku", "age": "12", "eye_color": "Green", "source": "ghibli"}

animedex ghibli person 267649ac-fb1b-11eb-9a03-0242ac130003 --jq '{name, gender, films}'

People project to Character via animedex.backends.ghibli.models.GhibliPerson.to_common().

Locations, Species, And Vehicles

The remaining snapshot families are exposed as rich, lossless models:

animedex ghibli locations --terrain forest --jq 'map(.name)'
animedex ghibli species --jq '[.[].name]'
# => ["Human", "Deer", "Spirit", "God", "Cat", "Totoro", "Dragon"]

animedex ghibli vehicles --jq '[.[].name]'
# => ["Air Destroyer Goliath", "Red Wing", "Sosuke's Boat"]

Endpoint Summary

Command	Python entry point	Purpose
`films`	`animedex.backends.ghibli.films()`	list film records with local filters
`film <film_id>`	`animedex.backends.ghibli.film()`	return one film by Studio Ghibli API UUID
`people [name]`	`animedex.backends.ghibli.people()`	list people with local filters
`person <person_id>`	`animedex.backends.ghibli.person()`	return one person by UUID
`locations`	`animedex.backends.ghibli.locations()`	list location records
`location <location_id>`	`animedex.backends.ghibli.location()`	return one location by UUID
`species`	`animedex.backends.ghibli.species()`	list species records
`species-by-id <species_id>`	`animedex.backends.ghibli.species_by_id()`	return one species by UUID
`vehicles`	`animedex.backends.ghibli.vehicles()`	list vehicle records
`vehicle <vehicle_id>`	`animedex.backends.ghibli.vehicle()`	return one vehicle by UUID

Gotchas

The high-level backend is intentionally offline. Use animedex api ghibli /films only when the user asks for live upstream data.
The public mirror is a frozen community data set. The snapshot reflects the mirror’s shape and contents at capture time; it is not a complete filmography database.
Source attribution still applies. TTY rows render [src: ghibli] and JSON includes the source tag metadata even though the data came from a local file.

The Python library page covers the same surface from inside Python.

animedex ghibli