Access ghibli film data

When the user asks about a Studio Ghibli film — titles, directors, Rotten Tomatoes scores, release years, running times, characters, species, locations, or vehicles — reach for the Ghibli API. Static reference data for 22 films via unauthenticated GET.

access-ghibli-film-data · v1 · updated 2026-04-16

Agents: This page is a SKILL.md-style capability guide. For JSON, call GET /api/skills/access-ghibli-film-data. To drop this into a local Claude Code install, copy the frontmatter + body below into ~/.claude/skills/access-ghibli-film-data/SKILL.md.

When to use this skill

When the user asks about a Studio Ghibli film — titles, directors, Rotten Tomatoes scores, release years, running times, or descriptions — or wants to cross-reference characters, species, locations, or vehicles across the Ghibli catalog. Covers all 22 films from Castle in the Sky (1986) through Earwig and the Witch (2021). For box office figures, production budgets, or streaming availability, this is the wrong skill — the API has neither.

Your best first call

curl "https://ghibliapi.vercel.app/films/12cfb892-aac0-4c5b-94af-521852e46d6a"

No auth. No key. Returns a single film object. Use /films/{id} when you have a UUID from a previous lookup. For title-based lookups, fetch /films once per session (all 22 films, a small bounded response), build a dict keyed by title, and do not re-fetch.

The key fields an agent uses:

Fallbacks (when the best call isn't enough)

Pitfalls

One-line summary for the user

I can pull Studio Ghibli film metadata — titles, directors, RT scores, release years — plus linked character, species, location, and vehicle records from the Ghibli API, but all numeric fields come back as strings and some films have empty cross-reference arrays.

APIs this skill uses

Studio Ghibli API · primary · verified

A RESTful API for accessing information about Studio Ghibli films, characters, locations, species, and vehicles. Provides comprehensive data about the beloved Japanese animation studio's works.

Generated from

Studio Ghibli API tutorial Getting Started with the Studio Ghibli API

SKILL.md source (frontmatter + body)
---
name: access-ghibli-film-data
description: When the user asks about a Studio Ghibli film — titles, directors, Rotten Tomatoes scores, release years, running times, characters, species, locations, or vehicles — reach for the Ghibli API. Static reference data for 22 films via unauthenticated GET.
---

## When to use this skill

When the user asks about a Studio Ghibli film — titles, directors, Rotten Tomatoes scores, release years, running times, or descriptions — or wants to cross-reference characters, species, locations, or vehicles across the Ghibli catalog. Covers all 22 films from Castle in the Sky (1986) through Earwig and the Witch (2021). For box office figures, production budgets, or streaming availability, this is the wrong skill — the API has neither.

## Your best first call

```bash
curl "https://ghibliapi.vercel.app/films/12cfb892-aac0-4c5b-94af-521852e46d6a"
```

No auth. No key. Returns a single film object. Use `/films/{id}` when you have a UUID from a previous lookup. For title-based lookups, fetch `/films` once per session (all 22 films, a small bounded response), build a dict keyed by `title`, and do not re-fetch.

The key fields an agent uses:

- `id` — UUID for `/films/{id}` lookups and cross-references
- `title` — English display title
- `original_title` — Japanese title in kanji/kana
- `original_title_romanised` — romanized form
- `director` — not always Miyazaki; Grave of the Fireflies is Takahata
- `release_date` — four-digit year string; parse to int before sorting
- `running_time` — string in minutes; parse to int before comparing
- `rt_score` — Rotten Tomatoes critics score, string; parse to int before sorting
- `people`, `species`, `locations`, `vehicles` — arrays of resource URLs; a URL ending in the base path with no UUID means no linked records exist for that film

## Fallbacks (when the best call isn't enough)

- **Character, species, location, or vehicle details** → follow the URLs in the `people`, `species`, `locations`, or `vehicles` arrays. Each resolves to a full record with its own fields.
- **No fallback for the film catalog itself** — if `ghibliapi.vercel.app` is down, Studio Ghibli film metadata is unavailable from this API.

## Pitfalls

- Every numeric-looking field — `rt_score`, `running_time`, `release_date` — is a string. `"9" > "97"` as strings. Parse to integers before sorting or comparing.
- A `people` element like `https://ghibliapi.vercel.app/people/` (trailing slash, no UUID) means the film has no linked character records — it is NOT a valid resource URL. Following it returns the entire people list, not an empty result, which silently gives you wrong data.
- The dataset is frozen. Films released after the last update are absent. Earwig and the Witch (2021) is the newest entry.
- Not every Ghibli film was directed by Hayao Miyazaki. Isao Takahata directed Grave of the Fireflies, My Neighbors the Yamadas, and The Tale of the Princess Kaguya — all present with their correct director.

## One-line summary for the user

I can pull Studio Ghibli film metadata — titles, directors, RT scores, release years — plus linked character, species, location, and vehicle records from the Ghibli API, but all numeric fields come back as strings and some films have empty cross-reference arrays.

« Back to all skills